ECON2206 Introductory Econometrics Week 5 Tutorials Readings • Read Chapter 4.1-4.2 thoroughly. Review Questions (these may or may not be discussed in tutorial classes) • What is the null hypothesis about a parameter? • What is a one-tailed (two-tailed) alternative hypothesis? • In testing hypotheses, what is a Type I (Type II) error? What is the level of significance? • The decision rule we use can be stated as “reject the null if the t-statistic exceeds the critical value”. How is the critical value determined? • Justify the statement “Given the observed test statistic, the p-value is the smallest significant level at which the null hypothesis would be rejected.” • What is the 90% confidence interval for a parameter? • When the level of confidence increases, how would the width of the confidence interval change (holding other things fixed)? • Try to convince yourself that the event “the 90% confidence interval covers a hypoth-esised value of the parameter” is the same as the event “the null of the parameter being the hypothesised value cannot be rejected in favour of the two-tailed alterna-tive at the 10% level of significance.” • How would you test hypotheses about a single linear combination of parameters? • What are restricted and unrestricted models? • How do you compute the F-statistic, given that you have SSRs? • What are general linear restrictions on parameters? Problem Set (these will be discussed in tutorial classes) • Ch4 Q2 - Compute the p-value of the test performed in part (iii) using the standard normal approximation to the student-t distribution. • Ch4 Q5 • Ch4 Q6
ECON2206 Introductory Econometrics Week 4 Tutorials • You will need STATA for this tutorial. If you are attending face2face tutorials is recommended you bring your laptop. For all tutorial you should have STATA installed or have the ability of accessing STATA trough the UNSW myAcess service. Please check if this is the case before the class. • Prior to attending create a work folder on yourlaptop, say F:IEwork, and copy the following files to that directory WAGE2. dta; hprice1. dta; Ch3 C2. do; Ch3 C6 . do. These files will be used in class. • If you do have access to Stata, then feel free to try the questions beforehand. Readings • Read Chapter 2.5, 3.1-3.2, 3.5 thoroughly. • Make sure that you know the meanings of the Key Terms at the end of chapter 3. Review Questions (these may or may not be discussed in tutorial classes) • What do we mean when we say “regress wage on educ and exper”? • Why and under what circumstance do we need to “control for” exper in the regres- sion model in order to quantify the efect of educ on wage? • What is the bias of an estimator? • What is the requirement of the ZCM assumption, in your own words? • How do you know that the OLS estimators are linear combination of the observations on the dependent variable? • What is Gauss-Markov theorem? • What is an endogenous explanatory variable? Exogenous explanatory variable? • What is multicollinearity and what is its likely efect on the OLS estimators? Problem Set (these will be discussed in tutorial classes) • Wooldridge Ch3 Q2 • Wooldridge Ch3 QC2 • Wooldridge Ch3 QC6 • As a continuation of Ch3 QC6, suppose you generated residuals from the regression in part (1), call this variable IQres. Next generate residuals from the regression in part (2), call this variable lwageres. Finally run a simple regression of lwageres on IQres. What would you expect to obtain as the estimated coefficient on IQres?
ECON2206: Introductory Econometrics Week 3 Tutorials • You will need STATA for this tutorial. If you are attending face2face tutorials is recommended you bring your laptop. For all tutorial you should have STATA installed or have the ability of accessing STATA trough the UNSW myAcess service. Please check if this is the case before the class. • Prior to attending create a work folder on your computer, say F:IEwork, and copy the following files to that directory lecture1. do; wage1. dta; lecture1alt. do; wage1 . raw; meap93. dta; meap93. des; Ch2 C6. do; charity. des; charity. dta and Ch2 C7 . do. You can find them all on Moodle in the Data and do-files section. Feel free to try the questions beforehand. If you save the file in another folder be sure you know in which folder is saved. Readings • Read Chapter 2 and Chapter 3.1-3.2 thoroughly. • Make sure that you know the meanings of the Key Terms at the end of Chapter 2. • Read handout “A brief introduction to Stata: 2016”. Review Questions (these may or may not be discussed in tutorial classes) • The minimum requirement for OLS to be carried out for the data set (xi , yi ), i = 1, . . . , n with the sample size n > 2 is that the sample variance of x is positive. In what circumstances is the sample variance of x zero? • The OLS estimation of the simple regression model has the following properties: — the sum of the residuals is zero; — the sample covariance of the residuals and x is zero. Why? How would you relate them to the “least squares” principle? • Which of the following models is (are) nonlinear model(s)? — sales = β0 /[1 + exp(-β1 ad expenditure)] + u; — sales = β0 + β1 log(ad expenditure) + u; — sales = β0 + β1 exp(ad expenditure) + u; — sales = exp(β0 + β1 ad expenditure + u). • Can you follow the proofs of Theorems 2.1-2.2? Problem Set (these will be discussed in tutorial classes) • Run lecture1 . do to reproduce the scatter plot and correlation between wages and education that were included in Lecture 1. Sometimes the data are not provided as a Stata data set. Run lecture1alt . do which does the same analysis after reading the data from wage1 . raw. Notice the diferences. • Wooldridge Ch2 QC6 • Wooldridge Ch2 QC7
ECON2102 – Macroeconomics 2 Additional Tutorial Questions Tutorial 2 Question 1 – Consider the Solow Swan Model whereby the economy is initially on its steady state. Suppose there is a permanent increase in s arising from the government increasing public savings. a) What is the effect of the increase in s on the steady state level of capital per worker? Illustrate using a diagram. b) Describe the likely effect(s) of a permanent increase in s on output per worker and capital per unit of labor. c) Draw graphs illustrating how the following change: - National savings rate - Growth rate of output per worker - Growth Rate of capital per worker - Capital per worker. You should draw these graphs as a function of time. Show how each variable was tracking before the permanent change and then after the permanent change. What is the difference between the savings rate effect compared with the effect of the remaining three parameters? d) What is likely to happen to consumption expenditure immediately after the rising savings rate? Will consumption expenditure over the long-run period exceed its level prior to the increase in s? Question 2 – Consider the basic Solow Swan Model. Suppose we have a Cobb-Douglas production function expressed as: f(k) = kα and s > n + g + δ. Describe in words, the likely effect of a rise in capital’s share of national income from α0 to α1. Draw a diagram to illustrate the effects on the steady state level of capital. Question 3 – Consider the Solow Swan model. Describe in words, the likely effect of a rise in worker effort or intensity. Draw a diagram to illustrate the effects on the steady state level of capital.
Module code and Title IFB315TC Supply Chain Modelling and Optimisation School Title School of Intelligent Finance and Business Assignment Title Case Study Submission Deadline 3rd November 2025 Final Word Count NA Students 1. Submission Deadline: 5 pm Monday 3rd November 2025. 2. Format Requirements: • Assignments must include the cover page with your student ID. • The format of the assignment comprises simply a response to each of the requirements. Thus, the assignment is not a report: Introductions, Executive Overviews, and Tables of Contents are NOT wanted. • All assignments must be typed, proof-read, and professional in appearance • The word font should be 'Arial' or 'Calibri' and font size 12 with 1.5 spacing. • The left and right margins should be 'Justified'. • Harvard Referencing if citation and references apply. • ALWAYS make variables italic or blue in paragraphs where discuss about them. 3. Submission Requirements: • Name Word file as IFB315TC-CW2-Your ID. • A group should submit one document a. Word format of Part 1 to LMO for group questions. Each student should also submit two documents, including b. Word format of Part 2, c. Vensim file of Part 2 to LMO for individual questions. Only electronic submission is accepted and no hard copy submission. • All students must download their file and check that it is viewable after submission. Documents may become corrupted during the uploading process (e.g. due to slow internet connections). However, students themselves are responsible for submitting a functional and correct file for assessments. Assessment tasks: Simulation of Supply and Demand of Filtration Devices in Dialysis Machines Using Vensim PLE Braun Melsungen AG is a German company that offers products and solutions for infusion therapy, nutrition therapy and pain management. The company safeguards the functionality of medical care that requires operations planning, including reliable and upgradeable supply chains to integrate smart IT solutions into the hospital processes. Braun Melsungen AG is also a leading company in the hemodialysis market. You are in charge of the supply and demand planning of filtration devices (let’s say filters) used in dialysis machines made by Braun Melsungen AG. Figure 1: dialysis machines (left) and filters (right) Braun Melsungen AG is mostly efficient regarding dialysis machines market. However, it is common to see the company experiences periods of low capacity utilization followed by substantial production increase and extra shifts. Your colleagues believe that the situation is due to the demand fluctuation and economy instability, but you have your own thoughts. Rechecking supply by your company and demand by hospitals over the last 120 months and drawing the following graphical statement about the corresponding supply and demand of filters, you get the impression that demand has more stability than supply. Figure 2: the supply by your company and demand by hospitals over the last 120 months Your aim is to determine what the reason behind this can be, and how you can resolve it. To do so, you want to simplify as much as possible your current situation to deal with it easier. The reason of simplification is to find solutions quickly in the line with a reliable supply chains and decide if you are on the right track. In addition, it is more effective to start with a basic model and extend it, than to create a complex model and attempt to obtain useful information from it after it is completed. Let us now take a few general observations about how the business operates. 1. The only source of supply is in-house production of filters in the company. 2. The only source of demand is sale to dialysis clinics in hospitals. 3. Inventory will never increase if there is no fabrication of filters. 4. The company cannot deliver filters to dialysis clinics in hospitals if there is not any inventory in its warehouse. 5. There is a current backup of 300,000 filters in the company stored in the case if there is no worker. 6. The process of training a newly-haired employee to be eligible for work takes two months under the most favorable conditions. 7. Similarly, the process of ending the contract of a currently employee also required a two months’ notice. 8. A trained employee makes 1,000 filters per month and under the best-case scenario in which he productively works the entire 8 hours a day. However, in reality, an employee is 6.4 hours productive a day. 9. Currently, 125 employees work in the company to fulfill a fixed demand of 100,000 filters per month. Part 1 Assume that manufacturing of filters is related in a make-to-stock (MTS) strategy with a “non- adoptive” manner. For such a manner, production is required to first fabricate filters and then sell. This is to say that the demands of filters are not immediately met when supply happens. 1. Your task is to draw a stock and flow diagram related to the non-adoptive supply and demand, and then insert its Vensim image in your Word document. (group, 15 Marks, 50 words limit) 2. There is a set of variables used in the stock and flow diagram. Briefly explain variables used in this diagram and relationship between them. (group, 20 Marks, 350 words limit) Part 2 3. It is now vital to realize the way supply of filters is decided. To do so, consider the following statement: “Investment and capacity are important and stable over a long-term vision. However, the company should hire or dismiss employees to a desirable point every 2 years once if there is a change in the demand.”. Let’s assume that monthly demands in the first, second, third, fourth, and fifth two-year time intervals are fixed and equal to 100,000, 120,000, 70,000, 110,000, and 110,000 units of filters, respectively. Your task is: a. to write down the Vensim equations used for this model with a setting of 120 months (add all of them to Word document) b. to draw strip graph figures of changes of stock levels and axillary variables developed in the model for the non-adoptive manner of supply and demand of filters (insert them in your Word document), c. to make a table depicting the sensitivity analysis for each of five two-year time intervals. (Individual, 30 Marks, 200 words limit) 4. Now, assume that manufacturing of filters is related in a make-to-stock (MTS) strategy with an “adoptive” manner. Due to the adaptivity, the stock decisions are also based on the current or recent demands. The stock correction that is the correction for a deviation of stock from its desired value due to unpredictability in demand of filters is not an immediate action. Instead, the stock correction can last two months. Similar to Q3, assume the monthly demands in the first, second, third, fourth, and fifth two-year time intervals are fixed and equal to 100,000, 120,000, 70,000, 110,000, and 110,000 units of filters, respectively. In addition, consider the following statement: “Because of poor demand forecasting in the first three two-year time intervals, overstocking equal to the triple of demand is unavoidable for each of these three two- year time intervals. However, the demand forecasting can be improved in the last two two-year time intervals and so there is no need to overstock filters in the corresponding two two-year time intervals.”. Your task is: a. to draw an extended stock and flow diagram related to the adoptive supply and demand, b. to write down only the additional Vensim equations used for this model with a setting of 120 months (add all of them to Word document) c. to draw strip graph figures of changes of stock levels and axillary variables developed in the new model (insert figures in your Word document) d. to make a table depicting the sensitivity analysis for each of five two-year time intervals. The Vensim file ended with PART2-Your ID.mdl should also be submitted. (Individual, 35 Marks, 250 words limit) Word limit: 850 words (not including data presented in figures or attachments). Marking Criteria Requirement Pts Performance indicator Part 1 Word: IFB315TC-CW2-Your Group ID.docx Question 1 15 Draw diagram correctly. A clean and readable diagram. Question 2 20 Good understanding of its variables and their relationships. Part 2 Word: IFB315TC-CW2-Your ID.docx Vensim: IFB315TC-CW2-Your ID.mdl Question 3 30 Draw graphs correctly. Good equation setting and the table for 10 years. Question 4 35 Draw diagram and graphs correctly. Good equation setting and the table for 10 years.
MIDTERM Assignment CHMS 5030: Molecular Analysis Part 1: Basic Concepts (50 pts) Q 1: Which wavelength (meter) could be used by Infra-Red (IR) spectrometer? (2 pts)_______. A) 10-10 m B) 10-9 m C) 10-7 m D) 10-5 m E) 10-3 m Q 2: A carbonyl C=O bond vibrates at 1710 cm-1 in IR spectra. What is the frequency in Hertz of light this carbonyl C=O absorbs (IR)? (2 pts)_______. A) 5130 ×1012 Hz B) 5130 ×109 Hz C) 5130 ×1010 Hz D) 5130 ×1011 Hz Q 3: Which vibration mode will give rise to the highest absorption frequency for a C-H bond in Infrared spectrum (IR)? (2 pts)_______. A) Symmetric stretching B) Asymmetric stretching C) Scissoring D) Twisting E) Rocking Q 4: If a carbonyl C=O bond vibrates at 1715 cm-1 in the IR spectrum. What is the overtone of this carbonyl (2 pts)_____. A) 3430 cm-1 B) 2230 cm-1 C) 1715 cm-1 D) 857 cm-1 Q 5: In the IR spectrum, which covalent bond has the lowest stretching frequency according to Hooke’s Law? (2 pts )___________. A) C-O. B) C-H. C) C-Br. D) C-N. Q 6: In the IR spectrum, the calculated O-H stretching wavenumber is 3553 cm-1 , what is the stretching frequency of O-D (Oxygen-deuterium bond)? (O-H and O-D have the same Force Constant) (2 pts )_____________. A) 3553 cm-1 B) 3053 cm-1 C) 2585 cm-1 D) 2085 cm-1 Q 7: In the IR spectrum, a carbonyl (C=O) stretching frequency was observed at 1685 cm-1 . Which type(s) carbonyl it might be______________? (2 point) Q 8: In the IR spectrum, a strong broad band near 3200 cm-1 is observed. Which of the following compound(s) is least likely to give rise to this band. (2 pts)_______. Q 9: In the IR spectrum, which of the following covalent bond(s) will have the lowest intensity of its stretching frequency_________? ( 2pts) A) C-O B) C-C C) C-H D) C-N. Q 10: In the IR spectrum, which of the following lactam carbonyl (C=O) vibrates at the highest stretching frequency _________? ( 2pts) Q 11: Which fragments or molecules could NOT be observed in the Mass Spectrum? (2 pts)________. Q 12: If a mass spectrum of morphine is requested to provide the molecular weight (MW 285), which of the following ionization method will be suitable (2 pts)____________. A) EI B) CI C) ESI D) MALDI Q 13: Which of the following reagents would cause less fragmentation when it is used in the chemical ionization (2 pts)____________. A) H2 (hydrogen) B) CH4 (methane) C) i-C4H10 (isobutane) D) NH3 (ammonia) Q 14: Calculate the degree of unsaturation for the following formula (2 pts). A) C17H19NO3: ________________________ B) C47H51NO14:_______________________ C) C7F5ClO:__________________________ D) C5H9BrMgSi:________________________ Q 15: In the MS spectrum (EI) for a small organic compound, it is found that the molecular ion peak is at m/z 285 with 85% relative intensity. The peak at m/z 286 has 15.8% relative intensity. Please propose a possible formula for this organic compound (2 pts)______. A) C20H29O B) C19H27NO C) C18H21O3 D) C17H19NO3 Q 16: In the MS spectrum (EI) for a small organic compound, it is found that the molecular ion peaks are at m/z 183 (100% relative intensity), m/z 184 (7.6% relative intensity), m/z 185 (97% relative intensity), and m/z 186 (7.4% relative intensity). Please propose a possible formula for this organic compound (2 pts)______. A) C7H5BrO B) C13H28 C) C7H14Cl2O D) C6H12Cl2ON Q 17: Use “Thirteen rule” to determine the molecular formula of an organic hydrocarbon compound (containing only H and C) with the molecular ion peak at m/z 100 in the Mass spectrum? (2 pts)____. A) C7H16 B) C8H4 C) C6H12 D) C7H14 Q 18: An organic compound is determined with high resolution mass spectrum that shows the molecular ion peak at m/z 285.1367 using electron ionization, which of the following formula is possible for this compound? (2 pts)____. A) C19H27NO m/z 285.2086 B) C18H21O3 m/z 285.1485 C) C17H19NO3 m/z 285.1365 D) C20H29O m/z 285.2211 Q 19: In the mass spectrum, the fragmentation process is usually complicated, but some general trends have been observed. Which of the following fragmentation pathway(s) will be generally favored (2 pts)_________. Q 20: Which of the following fragmentation can NOT occur in the MS using Electron Impact ionization? (2 pts)_________________. Q 21: Which of the following nucleus can be detected by NMR? (2 pts)_________________. A) 13C B) 12C C) 16O D) 17O E) 14N F) 15N Q 22: For a 600 MHz proton NMR machine, at what frequency that 13C will absorb? (2 pts)______. A) 100 MHz B) 150 MHz C) 200 MHz D) 250 MHz E) 300 MHz Q 23: An increase of scan number is the most common approach to increase the NMR sensitivity. If scan number increases 100 times, how many times of sensitivity can be increased (2 pts)_____. A) 100 B) 80 C) 50 D) 10 Q 24: A proton resonates at 2.5 ppm in a 400 MHz NMR machine, what frequency in Hertz of this proton is observed in a 600 MHz machine? (2 pts)_____. A) 1000 Hz B) 1200 Hz C) 1500 Hz D) 2000 Hz Q 25: In 1H-NMR spectrum, the aldehyde proton (-CHO) resonates at a very high frequency, ca. 10 ppm. Which of the following factor(s) contribute(s) to such high frequency (2 pts)_____. A) Strong hydrogen bonding B) High electronegativity of oxygen C) Diamagnetic anisotropy of π-bond D) Electron delocalization Part 2. Spectra Analysis (50 pts) Question 26: Each of the following IR spectra is associated with one of the compounds below. Identify the compound I-V associated with each spectrum A-E. Please indicate the key functional groups near the corresponding vibrating bands (10 points). Please provide rationalization details on the space aprovided or additional page. Question 27: A small molecule with formula of C4H9NO was determined by IR spectral analysis as provided below. Please propose a possible structure with rationalization (10 points). Question 28: The mass spectra of 4-heptanone was provided below using electron ionization method. Please draw fragmentation mechanisms to rationalize the fragments at m/z 99, 86, 71, and 43 (10 pts). Question 29: The mass spectra of cyclohexanone was provided below using electron ionization method. Please draw fragmentation mechanisms to rationalize the fragments at m/z 68, 55, and 42 (10 pts). Question 30: The 1H-NMR spectrum of 4-methoxybenzaldehyde was provided below. Please assign the proton signals to the structure and provide full rationalization (10 pts).
Film, Activity or Game Analysis- ARIN1001ATA GLANCE: WEEK TOPIC FILM/S TUTORIAL ACTIVITIES GAME/S Week 1 Introduction No classes Week 1 Week 2 (Imaginaries) Cyber imaginaries and Sci-Fi futures Interview with William Gibson (linkhere) Kitchen Carnage by Josephone Starrs (linkhere) Readings Discussion: Streeter, T. (2011) The Moment of Wired, Chapter 5 in The Net Effect: Romanticism, Capitalism, and the Internet. New York and London: New York University Press. Gibson, W. (1984) [excerpt pp 62- 83] Neuromancer. London: Harper Collins. Mind Map Production Activity (read morehere) Neuromancer In the tutorials this week we will play a 1988 video game calledNeuromancer Readmore about this gameand see if you can identify the ideas, characters and places borrowed from Gibson's book and those that have been added. Is this how you imagined Gibson's 'cyberspace' to look and feel? (linkhere) Week 3 (Imaginaries) Critical imaginaries and the Techlash ‘It’s Alive’ Frankenstein (linkhere) ‘Meet the Monster’ Frankenstein (linkhere) Young Frankenstein (linkhere) Mary Shelley’s Frankenstein (linkhere) Alvin and the Chipmunks meet Frankenstein (linkhere) Metropolis, 1927 (linkhere) Madonna, Express Yourself, 1989 (linkhere) Queen, Radio Gaga, 1982 (linkhere) Readings Discussion: Morozov, E. (2011). Introduction in The net delusion: How not to liberate the world. Penguin UK. Flew, T. (2019).Guarding the gatekeepers: Trust, truth and digital platforms, Griffith Review. Mind Map Production Activity (read more via themodule) No Game - Week 3. The Terminator, 1984 Gun Shop (linkhere) The Terminator, ‘Ill be back’ (linkhere) T800 Endoskeleton (linkhere) Black Mirror Episodes (Available via Netflix)
AMME5902 Computer Aided Manufacturing Final Project-Part A This part of the project consists of setting the manufacturing of the housing that will contain the Arduino Uno and the ultrasonic sensor (HC- SR04) that are used in the part B of the project. The following figure illustrates the design of the housing. Figure 1: a) Case of the Arduino and Ultrasonic Sensor. B) housing The CAD model of the housing is available in CANVAS. The chief aim of the project is to carry out the proposed tasks and arrive at comprehensive answers to the questions noted below. The tasks and the responses to the questions must be presented in a report. The report must include an introduction, contents page, all graphics, Excel spreadsheets, discussion, references and conclusions deemed necessary to convince management that you have a good understanding of the problem (or opportunity!) at hand (presentation: 2 marks) The following tasks must be carried out: 1. Analise the machining of the housing by using at least 2 different machining patterns (zig, spiral etc). Compare both machining patterns in terms of total toolpath length and total toolpath time. the machining parameters must be specified and justified. These parameters include tool dimensions, feed rates, and spindle speeds. (4 marks) 2. Using Process Engineering methods, analyse an area of machining of the housing (3 marks) 3. Using a combination of CAMWorks and UPG develop a Post Processor for the Mori Seiki NV4000 (3 marks) The following questions must be answered. 1. Which one of the three CNC machines noted overleaf would you recommend using if your company needed 1200 housings manufactured at the lowest possible price? (2 marks) 2. Which one of the three CNC machines noted overleaf would you recommend using if your company needed 1200 housings manufactured in the shortest possible time frame? (2 marks) 3. Production is scheduled to start on Monday the 3rd of November 2025. What is the first week day (Monday to Friday) that the 1200 housings will be ready? (consider any public holidays) You may work Saturdays but at a 50% overtime penalty. You may not work on Sundays. (2 marks) 4. If all your machinery is purchased new, what is your machinery worth at the end of the project? (2 marks) 5. Which method (i.e. cheapest or quickest) would you prefer to use in your factory and why? (Please discuss this point in great detail). (3 marks) 6. What would be the sale cost of one housing when you need to include a 140% profit ? (2 marks) Assumptions: • The stock pieces are made from Stainless Steel and are sized as per the maximum dimensions of the housing. • All machine cuts are done in one pass. • Plunge cuts are made at 2/3 of the feed rates that traversing cuts are made. • Rapid Traverse (G00) moves the tool at 6000mm/min for the Mori Seki, 2000mm/min for the Cincinatti Milacron and 1100mm/min for the TMCMill. Submission The submission consists of the following 3 files. 1) The Solidworks/CAMWorks file of the housing that correspond to task 1 (with any of the machining patterns used in task 1) . 2) a video of the CAMWorks simulation of task 1 (with any of the machining patterns used in the comparison) with a maximum duration of 2 minutes. 3) A written report that describe all the tasks and answers to the questions. Your report should be a good, clear presentation on A4 pages and PDF format. All sketches are to be included. The page limit for the report is 25 pages (without including references and appendices). You must upload the two files of the submission combined as a .zip file into CANVAS. The name of the file should be as per your group number. THIS ASSIGNMENT SHOULD TAKE EACH AVERAGE GROUP MEMBER APPROXIMATELY 15-20 HOURS TO COMPLETE Parameter Value Comment Time Cost (ex GST) Machine A Cincinatti Milacron -$85,000.00 Min Motor Capacity 1.5 kW In X,Y & Z Max Spindle Speed 3,5000 RPM Max Feed Rate (SS304) 390 mm/min In X,Y & Z Machine Charge Out -$165.00/hr Operator Charge Out -$165.00/hr Auto Pallet Changer 3.0 mins Machine B TMC Mill -$50,000.00 Min Motor Capacity 0.85 kW In X,Y & Z Max Spindle Speed 4,500 RPM Max Feed Rate (SS304) 75 mm/min In X,Y & Z Machine Charge Out -$98.00/hr Operator Charge Out -$113.00/hr Manual Pallet Changer 40 mins Machine C Mori Seiki -$345,000.00 Min Motor Capacity 1.6 kW x 2 In X and Y Min Motor Capacity 3.0 kW x 1 In Z Max Spindle Speed 12,000 RPM Max Feed Rate (SS304) 700 mm/min In X,Y & Z Machine Charge Out -$195.00/hr Operator Charge Out -$190.00/hr Auto Pallet Changer 1.95 min Hardware Stainless Steel (Grade 304) -$7.00/kg ADMINISTRATION Infrastructure costs 37.5 hrs / week Rent+Plant Maint. etc. 37.5hrs / week -$3,605.00/week Administration costs 37.5 hrs / week Labour 37.5hrs / week -$2,266.00/week Overtime Labour -50%/hr (up to 2hrs) Overtime Labour -75%/hr (2-4hrs) Depreciation All machines Simple -30%/year Federal Incentive Start up grant (once only) +$200,85.00 Utilities Elect, Water -$386.00 / week
MATH 215 FALL 2025 Homework Set 8: §15.6 - 15.9 Only some of the questions on this and other homework sets will be graded. Due November 3, no later than 11:59pm, submitted through Gradescope. You may work on these problems in groups (in fact, this is encouraged!), but you must submit your own set of solutions. Please neatly show your work! Question 1: Find the mass and center of mass of the tetrahedron bounded by the planes x = 0, y = 0, z = 0, and 3x + 4y + 2z = 12, if the density of the region is given by f(x, y, z) = 2 + x. Question 2: Let E be the region between the cylinders x2 + y2 = 1 and x2 + y2 = 16, above the xy-plane, and below the plane z = y + 4. Evaluate Question 3: Take a sphere of radius R with mass density proportional to the square of the distance from the origin in such a way that the maximum density is 9. Cut this sphere with two planes that intersect along a diameter at an angle of π/3. (This shape should look roughly like the segment of an orange.) Find the mass of this wedge of the sphere. Question 4: (a) Sketch the region of integration for the integral Rewrite this integral as an equivalent iterated integral with integration order dy dx dz. (b) Sketch the region of integration for the integral Rewrite this integral as an equivalent iterated integral with integration order dx dy dz. (This one is a bit more challenging than the first part.) Question 5: (a) Let E be the region in the first quadrant that is above the line y = x/3, below the line y = 3x, and between the curves defined by xy = 3 and xy = 27. Sketch the region E and evaluate (Hint: Try u = xy and v = y/x.) (Another Hint: This problem is suspiciously similar to Question 6 from the Fall 2023 Midterm 2, in case you need inspiration.) (b) Find where f (x, y) = 3y2 - 3xy - 4x2 and R is the quadrilateral with vertices (0, 2), (3, 0), (5, 4), and (2, 6). Hint: There may be a straightforward but tedious way to solve this problem as well as a faster, more subtle, way to solve this problem. (∗cough∗ see part (a) ∗cough∗) Question 6: A massive body E of constant density of mass equal to one generates a gravitational potential at a point (0, 0, a) given by (We have set the gravitational constant equal to one.) In this problem we take E to be a solid ball of radius R centered at the origin. (a) Compute V (0, 0, a), assuming that a > R) (i.e. the point is outside the ball). (b) Compute V (0, 0, a), assuming that 0 ≤ a
ELEN90066 Embedded Systems Design Project Report 1 Introduction The purpose of this report is to detail the design process and methods used in completing the Kobuki Obstacle Course Project. The report documents the design of a Finite State Machine (FSM) and its implementation on a Kobuki robot so that it can navigate an unknown obstacle course. Furthermore, the process used to test the system and the adaptions made leading into the final demonstration will be explained. Finally, the report will detail the outcome of this demonstration and a discussion of shortcomings and analysis of how the system could be improved. 1.1 Background Designing behaviour that allows a vehicle to traverse a complicated environment is an increasingly relevant problem. Finite State Machines provide a reliable method of achieving this, creating a way of representing the complicated behaviour in a readable format that the designer can follow. This representation of an output of the system in distinct ”states” also permits a more robust analysis and validation of the system, such as demonstrating that given a particular input, the system will always enter the desired state. This feature becomes critical for safety when working with systems that impact the physical world. The process and elements of the design outlined in this report could be applied to many real-world systems today. For example, the online retailer Amazon uses similarly-sized autonomous vehicles to transport racks of goods around its warehouses and fulfillment centres (E. Ackerman 2019). These vehicles would require similar (though far more complex) obstacle avoidance abilities as those developed in the Kobuki project. It is feasible to imagine an extension to the project that involves picking up an item at one end of the obstacle course and delivering it to the other. Moreover, full- size autonomous passenger vehicles use a similar process to operate on public roads today. Their objective is much the same as in the Kobuki Project: drive from point A to point B while avoiding obstacles in its path. The primary diferences lie in the complexity of the system, requiring the use of many more sensors and actuators, and the incredibly tight performance tolerances needed in such a high-risk environment. The Kobuki provided a platform for implementation and testing of our finite state machine. It is a relatively low-cost piece of equipment, ideally suited for education and experimentation. The hardware contains various sensors (discussed later), and was interfaced to the MyRio for processing. 1.2 Requirements The design requirements, provided in the project brief, gave an outline of the behaviour required of the Kobuki, though the exact algorithm and its implementation were left to be designed. The requirements for the Kobuki’s behaviour were: • The robot should not move until the ’Play’ button is pressed • The robot should maintain its initial orientation and drive forward in the absence of obstacles or a hill. • The robot should avoid obstacles, not fall of clifs or edges, and should maintain both wheels on the ground. • Avoidance must always be satisfied, even if multiple obstacles are encountered at once or in rapid succession. • After avoiding an obstacle the robot should return to initial orientation. • The robot should detect an incline and orient itself to drive straight uphill. • The robot should continue driving straight when the top of the incline is reached. • The robot should detect a downhill slope and orient itself to drive straight downhill. • The robot should detect level ground and stop within 40cm of the end of the descent. • The robot should not rotate more than 180。. • The robot should not exhibit 'hugging' or eratic behaviour. • The robot should not terrminate operation unexpectedly. • The course should be completed within the time limit. A sample course layout was also provided (Figure 1) in the project guide, although the exact course specifications were not given in advance. This obviously meant that a good level of robustness/ generality was required to ensure that our robot could complete any given course. Not knowing the exact course specifications also made a more thorough testing procedure essential in order to catch behavioural quirks and problems in edge cases. Figure 1: Sample course layout Knowing a more about how the course would be structed, some assumptions were made in the creation of the state machine algorithm. As the Kobuki was not required to avoid/detect any obstacles during hill climb and descent, the obstacle avoidance would only need to occur in the first section of the course. Similarly, there were no holes or drops in the obstacle avoidance section of the course, so wheel drop sensors were not necessary. With this understanding, we proceeded to design our Finite State Machine and incrementally adjusted it until it performed as required. 2 Finite State Machine From the understanding of the Kobuki’s behavioural requirements and the course structure, the design for our statechart was a hierarchical extended state machine, implemented using the Lab- VIEW statechart. This seemed to be the most logical approach, allowing the broader segments of the course to be compartmentalised, allowing for easier design and testing. LabVIEW provided advantages in the many aspects of the design, implementation, testing and validation. The graphi- cal nature of the LabVIEW statechart made it easier to implement the algorithm structure for the design and state refinements were simple to add. Furthermore, LabVIEW was much better suited during implementation, as it could communicate with the myRIO, which controls the behaviour of the Kobuki, through a wifi network to provide feedback on the state it was in and the values of the sensors. While the Kobuki calculates the distance it has driven and the angle it has turned by comparing the encoder values on each drive wheel, this is difficult to represent in the FSMs. In the following diagrams, states that depend on a distance driven or angle turned will use a variable, dist (mm) or angle (degrees), to store that current value with a default transition that increments the variable. This represents the behaviour of the Kobuki, though physical complications, such as the ramp angle or a high-friction surface, would change the rate that this increment actually occurs. Also note that while reading the FSMs, some of the transition actions had to be placed below the transition guard to condense the diagram so it could be read on an A4 page. Similarly, the true / guard was omitted from the default transitions to save room. 2.1 Hierarchical FSM Diagram The implementation of the state machine was done using the conventions described in An Introduc- tory Lab in Embedded and Cyber-Physical Systems v.1.70.. The top level FSM, shown in Figure 2, was a relatively simple extension on the skeleton statechart provided in the workshop documents. Adding a RESET state made testing the system much faster, as it would not require redeployment to test the same behaviour in a diferent scenario. Initially, we had hoped not to need to use the clif sensors, as the Kobuki would orient itself to point up the hill once it detected that it was no longer on flat ground. However, the performance became much more robust with the addition of the clif sensors as a fail-safe, in case the Kobuki did not detect the hill in time, or started to drift from its alignment. Due to the critical nature of its operation, the transition into the DETECT CLIFF state preempts the behaviour in the refinement of RUN, preventing the Kobuki from performing an action or transition that might cause it to drive of the clif. Figure 2: Top Level FSM The behaviour in the refinement of the DETECT CLIFF state, displayed in Figure 3, was relatively simple. When the sensors detected a clif, the Kobuki stopped and, depending on which sensor detected the clif, the opposite wheel reversed to turn away from the clif. When the 20。turn was complete, it would signal that the process was complete, so that it could transition back to RUN in the top level FSM. In this implementation, the centre clif sensor is not used. This is because the centre clif sensor does not provide information about the best way to turn to avoid the detected clif, as well as issues with it falsely detecting a clif when the Kobuki reaches the top of the ramp. Due to the location of the side clif sensors, one of them would always trigger before the Kobuki could drive of an edge, making the performance still robust without the centre sensor. Figure 3: FSM CLIFF DETECT The RUN state refinement, shown in Figure 4, handled almost all of the movement behaviour of the Kobuki. The Kobuki begins moving by entering the DRIVE state, where it simply moves directly forward until it detects an obstacle from a bump sensor or detects the ramp based on accelerometer values. An APPROACH state was also added between DRIVE and HILL CLIMB where, after detecting a hill, the Kobuki would drive forward its width of 30cm before transitioning. This also served to safeguard the behaviour from falsely detecting a hill due to noisy accelerometer values and being unable to transition back. In its current implementation, the R UN state does not allow the Kobuki to return to drive after transitioning to HILL CLIMB. This was decided based on assumptions about the course, knowing that the hill is the last obstacle. However, this does have an impact on the robustness of the behaviour and its applicability to a more generalised task, where this assumption might not be accurate. Figure 4: FSM RUN When the right or centre bump sensors were triggered while in the DRIVE state, the refinement would transition to LEFT AVOID, meaning that it would turn left to avoid the obstacle that was either on the right or directly in front. The FSM for the refinement of LEFT AVOID is shown in Figure 5, with the behaviour of RIGHT AVOID turning the opposite directions, but otherwise being identical. The obstacle avoidance behaviour is a little complicated, but can be better understood by following the arrows of the same colour through the diagram. Ideally, upon detecting an obstacle, the Kobuki would reverse, turn left, drive 30cm and turn right to face forward again, having avoided the obstacle. If during this drive it encounters another obstacle, it transitions to the green path, where it reverses, turns 180。, drives 50cm and turns to face forward again. If in this process, it again hits an object, the blue path will be taken and it will stop driving to immediately turn and face forward again. Figure 5: FSM AVOID Figure 6: FSM HILL CLIMB Finally, the last state refinement of RUN is the HILL CLIMB state, in Figure 6, showing the behaviour of the Kobuki on a hill and how it reaches the final STOP state. HILL CLIMB is centred around the orientation procedure, which is the intial state and is reachable from both CLIMB and DESCEND. This works by comparing the accelerometer value in the y-axis to the guards to determine which way the Kobuki should turn, which it does until the y-axis value is near zero. A value of zero indicates that the Kobuki is driving parallel with the slope of the hill. When the Kobuki detects the flat hilltop using the x-axis accelerometer values, it transitions to FLAT and again to DESCEND when it starts downhill. In that transition, it also asserts descending, causing the Kobuki to transition to STOP in Figure 4 when it again reaches flat ground. 2.2 Sensor Data The primary sensors utilised in our implementation were the 3-axis accelerometer, three bump sensors (left, right, and centre) and three clif sensors. An accelerometer outputs the acceleration along each axis of its reference frame. Using this infor- mation we can determine if the robot is oriented up the hill or not. If there is any acceleration in the y-axis as shown in Figure 9, clearly the Kobuki is oriented at an angle and hence a correction needs to be applied. In much the same manner, this system is used to detect the flat top of the course as well as the descent. During the design and testing phase, a low-pass filter was added to reduce noise in the accelerometer signals. The three bump sensors are the means by which the Kobuki can avoid obstacles. On physical impact, the bump sensor asserts its respective signal. Therefore the bump sensors (and further the clif sensors) are pure signals. When a bump sensor signal is present, the Kobuki will turn to avoid the obstacle. The clif sensor signals are used in the same way. The Kobuki passes the sensor signals to the MyRio using a serial connection. Command signals are passed to the Kobuki actuators the same way. Figure 7: Diagram of the Kobuki platform. Figure 7 shows the Kobuki hardware. The various sensors, ports, and buttons are labelled, as well as some unused functionality such as the LEDS. Figure 8: Orientation of the accelerometer axes in the MyRio The orientation of the 3-axis accelerometer in the MyRio is shown in Figure 8. The MyRio was mounted on the top of the Kobuki, attached with velcro and the serial connector attached. The forward direction of the Kobuki corresponded to the positive x-axis of the accelerometer (Figure 9, with z-axis not shown). Figure 9: Mounting of the MyRio on the Kobuki, with the x and y orientation shown 3 Design Procedure The design procedure followed several steps. Due to the nature of the requirements, we could develop the obstacle avoidance algorithm and the hill climbing algorithm in parallel. Once the algorithms were behaving predictably, we implemented a transition linking the two regions. 3.1 Obstacle Avoidance Naturally the first problem to solve is the obstacle avoidance. As seen in the state diagram, this is the first region the Kobuki enters. Major parameters that were adjusted here include: ● Distance to reverse upon making a collision. ● Angle to adjust after reversing. ● Distance to travel in the new heading. These parameters contribute to the behaviour of the robot and the obstacles that it can avoid. 3.2 Hill Climbing The hill climb solution required use of the accelerometer data. It was not necessary to compute the tilt of the Kobuki as the raw accelerometer values were enough to determine if it was facing uphill. When facing uphill there should be no observable acceleration in the y-axis, hence this was a trigger to make the Kobuki reorient itself. These triggers were defined using a threshold, hence noisy measurements could sometimes trigger these thresholds prematurely. Therefore a low-pass filter was introduced to smooth these measure- ments. This was only introduced after testing our simulated design on the actual platform. We found that there were a few key diferences in the simulation and reality that required adjustment. These are discussed later. 3.3 Transition Events Another key aspect to success of the Kobuki is the transitions between hill climb and avoidance functionalities. The importance of these cannot be overstated - many groups failed due to transitions not behaving as expected. The main transitions are described below. ● Transitioning from obstacle avoidance to hill climb at the base of the hill. ● Transitioning from hill climb to the flat peak of the climb. ● Transitioning from the flat peak to the descent of the hill. ● Transitioning from hill descent to stopping at the bottom of the hill. Table 1: Transition guards Each of these transitions depend on accelerometer values and the state of the Kobuki when these values are measured. A summary is shown (Table 1). These transitions are very important, as if one is triggers prematurely, the Kobuki will believe it is further through the course than it is. 4 Testing Procedure The testing procedure was a mix between simulation and controlled testing on the actual Kobuki. The computer simulation proved valuable in the initial testing of the avoidance algorithm. It was much faster to run a simulation than to deploy onto the hardware and run a physical test, however simulations were soon phased out as there were key diferences between simulation and physical testing. A main contributor was the coordinate system between the simulation and actual platform were diferent (the y and x axis were switched) which became tedious to switch on each test. This meant that every transition guard using accelerometer values had to be adjusted to reflect the coor- dinate system between the simulation and physical system. The next was the introduction of noise - noisy measurements from the real system tended to cause the Kobuki to behave unpredictably, so testing in the simulation was less useful to rectify these problems. This meant the simulation was less useful for testing the hill climb section. The simulation did still provide value in developing the general approach to climbing the hill. Fine-tuning to develop precise accelerometer guards could only be done on the Kobuki hardware. While addition of the low pass filter solved some issues, the Kobuki motion itself occasionally caused peaks in the signal that were not filtered enough. It was found that when switching states involved a wheel speed alteration, high accelerations would occur, creating unpredictable results. Further, when the Kobuki collided with obstacles accelerometer peaks were also observed. This led us to introducing a delay between states which in turn allowed accelerometer signal transients to settle before the state diagram began making decisions again. This delay was made to be significant due to the importance of these transitions, outlined in the previous section. For the purposes of simplicity and readability, these delays were not represented in the FSM diagrams shown previously. Utilisation of the wifi feature on the MyRio was immensely valuable during testing. This allowed a real-time view of the sensor data provided by the Kobuki (displayed on the PC in LabView), as well as the current state.The real-time sensor data was instrumental in discovering the source of many unwanted transitions. It was possible to see which sensors were activated, and exact accelerometer output. Debugging time was thus vastly reduced, as when a problem occurred it was simple to identify the state the system was in and which transition should or should not be taking place. The easy debugging was especially useful in solving the accelerometer issue mentioned above, as when the physical Kobuki displayed unintended behaviour it was quick to find out that it was entering the HILL CLIMB state when it was not on a hill. The addition of the button B1 as a reset button was particularly useful in testing. This allowed the Kobuki to reset from its initial state after a test, without needing the code redeployed. Once again, this improved debugging time, allowing multiple variables to be examined in a single test. When the intended design was implemented and performed as would be expected, more rigorous testing of the system in scenarios similar to the course was undertaken. Whereas previously small tests were run to check the performance of one part of the statechart, such as avoiding a single obstacle or being able to orient up the hill, it was important to check how these behaviours worked together. This was intended to test our assumptions in the design of our algorithm, rather than to see if it had been implemented correctly. We had limited access to the actual obstacles and ramp that would be used in the final assessment. This meant that our system needed to be sufficiently general in order to cope with a variety of obstacles, and perform as expected on slopes of varying inclines, width, and surface texture. However, using the equipment available small sections of the course were set up to see if the Kobuki could reach the ramp after a few obstacles. The obstacle avoidance had trouble passing multiple obstacles, as if it became caught in a corner it would be trapped. Thus, additions were made to the avoidance algorithm to account for running into an obstacle while already avoiding another one, by turning around and attempting to pass in a diferent section. With these more extended tests, other issues were also noticed. Some Kobukis had a noticeable drift to them, and while intending to drive straight they would veer to one side. While it would be possible to correct for this by adjusting wheel speeds slightly, the error was diferent in each physical Kobuki so this adjustment would be largely pointless. Furthermore, each collision with an obstacle caused the Kobuki to lose its sense of the forward direction slightly, as the impact would create some deflection. This could not be entirely eliminated, but reducing the speed of the system when in DRIVE helped somewhat, which was helped by having more time to complete the course. The testing procedure involved repeated trial and error, particularly in determining the ideal ac- celerometer thresholds. This seemed to be the most efficient way to find robust thresholds, although at times proved tedious. 5 Validation Procedure To validate our algorithm we can notice that all states are reachable and no contradictions are possible. We are exclusively in hill climb or obstacle avoid and the transitions are very explicit. Examples that ensure correctness of the algorithm include: ● We only enter clif avoid when a clif sensor is activated. ● The pause state is only entered when button B0 is pressed. ● In the run state we exclusively avoid obstacles until an accelerometer event is triggered. To further validate the algorithm we can consider some potential unsafe states and prove that it is impossible to enter these states. To do this we need to define what an unsafe state in this context is. We use the following definition: If the Kobuki enters a state in which it is possible for the device to fall of a clif, or enters a loop in which it is impossible for the state machine to halt in the final Stop state, then the state machine is unsafe. The implemented state machine is not a composed state machine - it does not operate at the same time as another while occupying two states. Hence the Kobuki algorithm only occupies a single state at a time, greatly reducing the risk of dangerous states. An obvious unsafe state would be to be in OBSTACLE AVOID state when the Kobuki is on a hill, as we do not expect there to be clifs to avoid during the obstacle avoid section. However, this contingency was covered by CLIFF AVOID being a top level state. Meaning regardless of the state, we will always avoid a clif. With this feature the only remaining unsafe state will be if we cannot finally halt the state machine. Due to the linear nature of the state machine we are guaranteed to reach this state eventually assuming the appropriate string of accelerometer events. Hence we can conclude that the given state machine is safe. The design is restricted to this exact use case as it assumes the progression of the obstacle course. We assume the Kobuki first needs to avoid obstacles, and then climb a hill. It can be noticed that when the Kobuki has entered the hill climb state, obstacle avoid becomes unreachable. Further, once the robot completes the obstacle course and stops, it will never leave the stopped state (unless the reset button is used). While this achieves the specifications of the project, if the system needed to be generalised the state diagram will have to be updated. 6 Demonstration Outcome 6.1 Attempt One We were fortunate to have two attempts at the course, as our first one failed due to an anomaly in the course design. While the right hand side of the path was largely trivial, the left hand side presented a challenge in a small, diagonal pathway through which the robot barely fit. Figure 10 shows the small space that caused the Kobuki to fail. Figure 10: The Kobuki unable to pass through too small a gap Upon colliding with the obstacle, the Kobuki was then largely stuck as the avoidance algorithm first caused the Kobuki to reverse away from the obstacle. The Kobuki would then reverse into the wall immediately behind it, and as there are no collision sensors on the back of the robot, begin to lose it’s heading and fail the task. The robot was unable to manoeuvre successfully when the obstacles were too close together. 6.2 Attempt Two Our second attempt almost fell to the same issue. This time the Kobuki traversed the right hand side, but unluckily just clipped an obstacle, causing it to run into a wall. After colliding with the obstacle and the nearby wall repeatedly for some time, while making little progress, the Kobuki regained its heading and made it to the hill climb section. This section worked without fault, even displaying its clif sensing ability. The Kobuki climbed the hill while re-orienting to maintain a straight uphill heading, travelled along the plateau, and descended without issue. The overall time taken to complete the task was slightly beyond the time limit. The time constraint would have easily been met if not for the repeated collision issue earlier in the test. Figure 11 shows the Kobuki upon completing the descent and coming to a stop. Figure 11: Kobuki at the end zone Of the initial requirements, the following were satisfied: The robot should not move until the ’Play’ button is pressed The robot should maintain its initial orientation and drive forward in the absence of obstacles or a hill. The robot should avoid obstacles, not fall of clifs or edges, and should maintain both wheels on the ground. Avoidance must always be satisfied, even if multiple obstacles are encountered at once or in rapid succession. After avoiding an obstacle the robot should return to initial orientation. The robot should detect an incline and orient itself to drive straight uphill. The robot should continue driving straight when the top of the incline is reached. The robot should detect a downhill slope and orient itself to drive straight downhill. The robot should detect level ground and stop within 40cm of the end of the descent. The robot should not rotate more than 180◦ . The robot should not exhibit ’hugging’ or eratic behaviour. The robot should not terrminate operation unexpectedly. The course should be completed within the time limit.
Department of Electrical and Electronic Engineering ELEN90066 Embedded System Design Kobuki Obstacle Course Project Overview The first SIX workshops (Workshops 1—6) for the subject comprised a structured sequence of laboratory exercises designed to give you experience and develop competency in a real-life embedded system, a Kobuki robot platform, and the tools used to program it. The next THREE workshop sessions (Workshops 7—9) are devoted to optimising the design of the Kobuki in order to navigate an obstacle course in the final workshop in Week 12 of the semester. There are no in-class tasks to be performed so it is up to your team to determine how to best spend your workshop time. You will be writing a team report that documents your design and testing procedure so it is important that you document your design, implementation and testing procedure as you go. Your Kobuki obstacle course algorithm MUST be coded as a Finite State Machine (FSM) in either C or LabView. Obstacle Course Design Requirements The “B0” button on the Kobuki is referred to as the “Play” button. For CyberSim, a separate “Play” button has been added on top of the Kobuki top view which maps to the “B0” button. The following are the design requirements for the obstacle course navigation of the robot. 1. Startup: When powered on or reset, the robot shall not move until the “Play” button is pressed. 2. Run: The robot shall begin movement the first time the “Play” button is pressed and continue running until paused or stopped. (a) Ground Orientation: The ground orientation of your robot is the direction the front of the robot is pointing when it first runs. The ground orientation does not change after subsequent pausing/resuming; only after a power cycle, reprogram or restart of the robot or its embedded controller. (b) Drive: Your robot shall maintain ground orientation and drive forward while on level ground and clear of obstacles. (c) Obstacle Avoidance: Your robot shall avoid obstacles. i. Cliff Avoidance: Your robot shall not fall off of cliffs or edges. ii. Wheel Hazard Avoidance: Your robot shall avoid one or more wheels losing contact with the ground. If a wheel loses contact with the ground, your robot shall attempt to recover and move around the incident hazard. iii. Object Avoidance: Your robot shall avoid objects in its path. It is acceptable for your robot to touch objects as long as it immediately changes course in an attempt to avoid. iv. Avoidance Robustness: Obstacle avoidance must always be satisfied, even if multiple obstacles are encountered simultaneously or in short succession. v. Reorientaton: After avoiding an obstacle, your robot shall eventually reorient to ground orientation. (d) Hill Climb: Your robot shall have a Hill Climb ability. i. Hill Climb: When an incline is encountered, your robot shall drive uphill towards the top of the incline. It must not go over any edge of the incline. ii. Hill Plateau: When the top of an incline is reached your robot shall remain driving forwards along the plateau (top flat section) of the hill. iii. Hill Descend: When an downward slope is encountered, your robot shall drive downhill towards the bottom of the slope. It must not go over any edge of the slope. iv. Ground Stop: After climbing and descending, when the bottom of the slope is reached, your robot shall stop and terminate execution within a set distance of the bottom (40cm) as its final goal has been achieved. 3. Pause/Resume: The “Play” button on the top of the robot shall start/resume or pause movement of the robot. At any point the robot is moving, the “Play” button shall cause it to immediately and completely stop. Subsequently pressing the “Play” button shall cause the robot to resume operation. 4. Performance: When moving, your robot must satisfy the following performance char- acteristics: (a) Turnabout: Your robot shall never rotate in place more than 180◦ . (b) Chattering: Your robot shall not move erratically or exhibit chattering. (c) Abnormal Termination: Your robot shall not abnormally terminate execution, with the exception of power or mechanical failure. (d) Obstacle Hugging: Your robot shall not repeatedly encounter (“hug”) an obstacle for the purpose of navigation. (e) Timeliness: Your robot shall achieve its goals in a timely manner. Hint: Your robot does not need to follow a specific path or trajectory, or return to a specific path or trajectory following obstacle avoidance — it need only eventually return to and maintain its original ground orientation. Note: The Hill Climb section of the course will NOT have any obstacles present. Workshop 10 (Week 12) - Robot Obstacle Course In your final scheduled workshop for the subject in Week 12, your team will be running the Kobuki robot on an obstacle course in order to determine if your navigation and hill climbing algorithm is successful. The following subsections will provide details on the course set up, workshop schedule and logistics for the final workshop. Figure 1: Obstacle course layout. Note that items are not to scale. The course The course will consist of an area approximately 5m ×3m that will contain the following: • Barriers (walls) that mark the edges of the course. These may be bumped into by the Kobuki but not “hugged”; • Several shaped obstacles (e.g. rectangular, triangular) that will be in random positions on the course; • A hill section that comprises of an upward sloping incline, a flat top, and a downward sloping ramp. The Kobuki is not to drive off the side of the hill. An example course layout is shown in Figure 1. Note that this course does NOT correspond to the precise layout that your robot will be tested on — you will not know the locations of the objects on the course before you run your robot on it. The only information known about the course is that it will narrow towards the hill, which will be at the end of the course as shown in Figure 1. Your robot MUST stop within 40cm of reaching the bottom of the far side of the hill. Course location The obstacle course will be either located in the Telstra Creator Space Test Bed or the workshop locations PAR-173-L1-124-EDS 1. You will NOT be able to access the course until your final workshop. NO testing may be done on the actual course before the trial, even during class time. You team must make sure that you have simulated and tested the robot enough that you are confident it can handle any configuration of the course. Week 12 Workshop Logistics The class schedule for the final workshop will be as follows: • 0:00 - arrive at normal workshop room; • 0:05 - robots handed out to teams; • 0:10 - demonstrators will check the myRIO file systems to ensure all prior code is deleted; • 0:15 - myRIOs are to be programmed by teams. Once programmed, robots are to left ON and in the PAUSED state; • 0:30 - class heads to arena to run the robots on the course. NO alterations to the code are allowed from the moment the class leaves the workshop room; • 1:30 - class returns robots and all equipment. At the obstacle course, the robots will be tested with the following procedure: • A team will be selected at random; • The team will place their robot in the marked starting position, which is referred to as the ground orientation; • The demonstrator will signal when to press the “Play” button on the Kobuki and will start a timer once the button has been pressed; • The attempt is considered complete if any of the following occur: 1. the robot completes the course by successfully avoiding all obstacles, ascending and descending the hill and then stopping within 40cm of the bottom of the downwards sloping section; 2. the robot breaks any of the rules in the ‘Obstacle Course Design Requirements’ section; 3. the elapsed time is greater than 120 seconds; 4. the robot is interfered with in any way. Assessment (20%) The assessment for the Kobuki Project is worth 20% of the final mark for the subject. Of this, the report itself is worth 18% and the project outcome worth 2%. Project Outcome (2%) Your team will be graded on how successful the robot was in achieving the objective of navigat- ing the obstacle course. The marking rubric for this component is integrated into the marking rubric for the report on LMS. Project Final Report (18%) The report must contain (but is not limited to): • a complete FSM diagram (including pause states) of the robot algorithm using the nota- tion covered in the lectures. If you use an Extended state machine, or state refinements, you must show all states, variables and transitions. • a basic description of the sensor data available to the myRIO and how it was used in your algorithm. • your design procedure (this could include work from earlier workshops). • your testing procedure (this could include simulation data or screenshots). • your validation procedure (this could include reliability or reachability analyses). • the outcome of the run in the final workshop. Make sure to record this in the final workshop. You do not get a second chance at this. • a discussion section (how you applied knowledge from the lectures, what you would im- prove in the future). The report is to be no more than 20 pages in length. A marking rubric for the report is available on LMS. Submission Details Requirements for submitting the report: • Submission is via the LMS. • Submissions must be in PDF format. • ONE submission per team. • The report is to be no more than 20 pages in length, not including any appendices. • Late submissions will be penalised at the rate of 10% per day. • Submission date: Monday Nov 3 at 23:59.
This assignment is the last part of the individual project for which students will use a vendor system to address a risk problem of their choice, implementing an analytical model in Beacon or risk assessment in Archer. Preliminary: Basic exercise in Archer or Beacon (10 points) Part 1: Problem statement and Business Requirements (10 points) Part 2: Implementation and presentation of the following: (80 points) ● Option 1 - Operational Risk Assessment in Archer ○ Create Dashboard ○ Build Risk Assessment Questionnaire ○ Populate data Produce reports and analysis ● Option 2 - Financial Risk Model in Beacon ○ Develop a new instrument or extend an existing one using Beacon IDE ○ Create some positions in a book ○ Capture some risk analytics on the book of positions ○ Produce reports and analysis Details For either choice of implementation, the following artifacts are expected: ● Problem definition and approach (word document) (DRAFT and FINAL) ○ Selection of a business (real or hypothetical) and definition of a context, including key stakeholders ○ Clear and detailed description of the risk problem ○ Your approach / methodology and project deliverables, and how they address the risk problem ● Implementation (code / configuration in Beacon / Archer) ○ Implementation consistent with style. of examples ○ Clear code / configuration logic ○ Logical user interface and clear reports/outputs ○ Works as specified ● Presentation of reports and results (PowerPoint) ○ Description of data captured ○ Clean and intuitive reports ○ Clear statement of risk problem ○ Quantitative assessment ○ Qualitative assessment of the results - explain how the stated problem has been addressed ○ Evaluation of the risk and potential mitigants Assessment The project will be evaluated on the following factors: ● Demonstrated understanding of the Risk Problem and how the implemented solution addresses it ● Demonstrated understanding of the Tool and solution implementation ● Artifacts follow best practices (professional, clear, concise) To guide you further, please refer to the detailed rubric below for a more nuanced breakdown of the evaluation criteria. Submission To complete your submission, 1. Click the blue Submit Assignment button at the top of this page. 2. Click the Choose File button, and locate your submission. 3. Feel free to include a comment with your submission. 4. Finally, click the blue Submit Assignment button.
BS3082 Cardiovascular and Renal Precision Medicine: Assessment instructions Oral Poster Presentation Assessment Design and present a poster outlining a cardiovascular or renal biomarker and its importance in the diagnosis and treatment of disease Precision medicine relies on validated biomarkers with which to better classify patients by their probable disease risk, prognosis and/or response to treatment. In this module, you will learn about important biomarkers in cardiovascular and renal diseases. For this assessment, you will design and present a poster outlining one cardiovascular or a renal biomarker (or one that has relevance in both contexts). You should use peer-reviewed evidence to outline its importance in the diagnosis and treatment of disease. You can present as many studies as is relevant (although ideally we would expect a minimum of five). This poster assessment mark accounts for 25% of the overall module mark for BS3082. Instructions: Your posters should be prepared (ideally using the PowerPoint template on Blackboard) in portrait. You will have five minutes to present your poster followed by two minutes of questions. What you should include in your poster: · An appropriate title · Introduction o An introduction to your chosen biomarker and associated disease o How can the biomarker be used to assess the presence and/or prognosis of the disease? o How can the biomarker be used to guide the treatment and management of the disease? · Methods o How did you identify the studies used in your poster? Did you hand search relevant articles? Did you take a more systematic approach using search terms to search relevant databases? · Results o Use relevant findings from research papers to inform. your poster and reference them on your poster o Please add at least one figure to demonstrate that importance of the biomarker o If you use a figure from a published manuscript. it needs to be referenced. Alternatively, you can use published data to create your own figure. · Conclusions o Summarise your information on your chosen biomarker o Are there any weaknesses of this biomarker? · Suggestions for future research · Reference list: Please use Vancouver reference formatting. Please add your name and student ID to the bottom right of the poster Please consider the following when designing your poster: · Keep your title short and to the point so as to grab the attention of the viewer · It is very important that the text and data included in your poster follow a logical and hierarchical order. When we approach new information we tend to read from top to bottom and from left to right. Therefore, it makes good sense to lay your work out in this order · Try not to present long and detailed sections of text. Bullet points (or other ways of separating text such as titles or headlines) can often be more effective and will maintain the reader's interest · Keep it simple, clear and concise. The poster needs to be eye-catching and attractive, but filling up your poster space with excess clutter can be distracting for the viewer. Ensure that all sections of text, visual images and figures are aligned correctly, and avoid big areas of space, this will ensure that the poster is well presented. Remember the poster should make a good visual impression! · Ensure that your font size is large enough to be legible from at least one to two metres away. Individuals will soon tire of having to lean in or squint to read small text. It is best to use fonts that are easy on the eye, such as Arial, Calibri or Times New Roman. · Maintain a consistent and clean style. throughout. Try to avoid bright, noisy or colours that clash · If used in the correct way, figures and visual images can greatly enhance your poster, increasing both understanding and interest. However, ensure that all graphics are relevant to your work, and linked by references e.g. figures numbers in the text. · Make sure that all diagrams are clearly captioned and easily seen. Captions should be positioned below to the diagram/figure and the figure should be placed close to the relevant text. · Please use in text citations and include these references in the reference list section Assessment set: Thursday 16th October 2025 Submission Deadline: 12pm noon, Thursday 6th November 2025 Poster Presentations: 1pm, Thursday 13th November 2025. Note: The oral poster presentations will be conducted in a conference-style. format. All students are required to stand by their posters for the entire session. Assessors will circulate, inviting you to present your work and answer questions. Please note that you must remain with your poster for the full duration of the session, even after you have been assessed, to engage with visitors viewing posters and support fellow students who are still awaiting assessment. Please use the marking criteria below to guide you with the requirements for this assessment Marking Criteria Layout, Design and Content (67%) 100, 95, 90, 85 Outstanding presentation of information and extremely clear. It is very easy to navigate around the poster. There is an exceptional range of relevant evidence presented. All relevant sections included. Demonstrated an exceptional understanding of findings. 80, 75, 72 The information is very well handled and presented. It is easy to navigate around the poster. There is a very good range of relevant evidence presented. All relevant sections included. Demonstrated a very good understanding of findings. 68, 65, 62 The information is well presented and clear. It is generally easy to navigate around the poster. There is a good range of relevant evidence presented. Most of the relevant sections included. Demonstrated a good understanding of findings. 58, 55, 52 The information is reasonably presented and clear. It is fairly easy to navigate around the poster in most areas. There is a satisfactory range of relevant evidence presented. Most of the relevant sections included. Generally demonstrated a solid understanding of findings. 48, 45, 42 Basic presentation of information, not always clear. It is challenging to navigate around the poster. Some of the relevant sections included. Unsatisfactory range of evidence presented. Poor understanding of findings. 38, 35, 30 Limited presentation of information, unclear in large areas. Very challenging to navigate around the poster. Poor range of evidence presented. Few of the relevant sections included. Very little understanding of findings. 25, 20, 15 Hardly any presentation of information, very unclear. Almost impossible to navigate around the poster. Hardly any evidence presented. No understanding of findings. 10, 5, 0 Impossible to navigate around the poster. No evidence presented or understanding of findings. Oral Presentation (33%) 100, 95, 90, 85 Outstanding presentation, demonstrating excellent knowledge beyond expectations, with very clear explanation of poster contents. Students made eye contact and engaged with the audience. Avoided reading off the poster. Student showed outstanding enthusiasm for the topic and answers to questions demonstrated a high degree of background knowledge. 80, 75, 72 Excellent presentation, demonstrating very good knowledge and explaining all the information clearly, students made eye contact and engaged with the audience. Avoided reading off the poster. Student showed enthusiasm for the topic and gave very clear, comprehensive answers to questions. 68, 65, 62 Very good presentation, demonstrating key knowledge with no important omissions, very good summary of key points, students made eye contact and engaged with the audience. Good interaction with audience and avoided reading off the poster. Student showed enthusiasm for the topic and answered the questions well. 58, 55, 52 Good presentation, demonstrating relevant knowledge but with areas that could be improved, summarising poster well but with some omissions, students mostly made eye contact. Mostly avoided reading off the poster. Student showed some enthusiasm for the topic and answered the questions adequately but with some gaps in knowledge or clarity. 48, 45, 42 Presentation was lacking in some areas, including knowledge/missed relevant parts of the poster, but achieved minimal expectations. Students made some eye contact and engaged with the audience. Read off the poster for the majority. Enthusiasm and answers to questions were adequate. 38, 35, 30 Presentation was lacking in clarity, demonstrating very little knowledge, not getting the message across, students made no eye contact and rarely engaged with the audience. Read off the poster and answers to the questions were poor. 25, 20, 15 Presentation was seriously lacking, demonstrating no knowledge, no evidence of engagement with audience. Read off the poster and could not answer questions. 10, 5, 0 Impossible to comprehend presentation. Completely lacking in all areas.
AMA1D04 Understanding Social Conflicts by Game Theory 2025/26 S1 Project instructions: Topic Free choice of topic on voting system, decision and strategy making, simulation of social situation by game theory, etc… the topic should be related to the society or student’s own discipline of study with application of the knowledge learnt from this course. First-come-first-served: in case too many students choose a similar topic, students who submit the project proposal later might be banned. The project proposal should be completed by using the form. provided and submitted onto Blackboard. Only after your proposal is approved you can work with the chosen topic. Format - English should be used. Names of the original language can be shown in brackets after the English translation. - Microsoft Word “.docx” file with normal margin, vertical A4 size, font size 12 and 1.5 line spacing, Times New Romans. - The file name should be “(your family name)_(your given name)_(proposal/project).docx”. For example, your proposal is “chan_taiman_proposal.docx” and your project to submit at the end is “chan_taiman_project.docx”. - The proposal should be included in front of the final project submission. The project body, excluding the proposal in front and the references or appendix at the back, should be within 1500-2500 words (figures, equations, tables not counted). Reference should be shown in APA format. If necessary, tables, coding or other additional information can be attached as appendix. - The final work should be submitted via Blackboard. Turnitin is used to check for similarity and AI reliance. Plagiarism is strictly forbidden. Over-reliance on AI will result in low score. Deadline Deadline of submitting the project proposal: 06/11/2025 23:59 (week 10 Thu) Deadline of submitting the completed project: 06/12/2025 23:59 (week 14 Sat)(tentative) Policy on the use of AI To implement the use of generative artificial intelligence (GenAI) in academic writing and to enhance digital literacy, students are encouraged to make use of GenAI tools in the assessment. Such tools include but not limited to Chat-GPT4, Microsoft Copilot, Google Gemini, etc. A list of GenAI tools can be accessed via platforms like genai.polyu.edu.hk. Before using GenAI in writing the assessment, you should read the following in advance: (i) University’s disciplinary regulations concerning conduct in examinations and, in particular, the regulations on academic integrity. (pp. 37–40, pp. 74–76) https://www.polyu.edu.hk/ar/docdrive/polyu-students/student-handbook/Student_Handbook_2024-25_English.pdf (ii) University’s “Guidelines for Students on the Use of Generative Artificial Intelligence”. https://www.polyu.edu.hk/ar/docdrive/polyu-students/Student-guide-on-the-use-GenAI.pdf We will introduce some techniques about using generative AI in project writing on the AI seminars during tutorial sessions. Self-learning materials are to be uploaded onto Blackboard. Report about use of AI If you opt to use GenAI in the assessment, you are required to submit an extra report for using GenAI in producing your work. On this report, you should explain how you use GenAI to assist you in the following tasks, and your rationale in adopting AI generated results: (1) How do you use GenAI to suggest project ideas and search for relevant reference? How do you narrow down the scope and decide a suitable topic? (2) How GenAI help you to write a proposal and structure for the assessment? Is there any amendment you made in the final outcome compared with the proposal/structure? (3) Which part of your project contain components in various forms (writing, mathematical evaluations, graphs and animation, data analysis, program coding, etc) created by GenAI? Why do you adopt or reject the work created by GenAI? Please justify. You can also show how you correct or modify the work done by AI so that it can be adapted in your project. (4) How GenAI help with enhancing the clarity and writing style. of your project? For these questions, you should mention GenAI tools used for each part clearly. You can also include some screen-capture to show your input commands (prompts) and the key results created by GenAI, no matter they are adopted in the final project or not. The report should be precise and within 1000 words, not counted towards your final project.
New Assignment 32025 Part 1 Select an appropriate model to train the dataset and make predictions (3 Points) The UCI Adult dataset-sometimes called the Census Income dataset-is a classic resource in machine learning for demonstrating classification tasks, particularly binary classification. Dataset Description ·Number of Instances:Around 48,842 rows(depending on whether duplicates/missing rows are handled). ·Number of Attributes:14 features(plus the target) ·Feature Types: ■ Numeric(e.g.,age,hours-per-week,capital-gain). ■ Categorical (e.g.,workclass,marital-status,occupation,sex). ·Target Column: ■ Labeled as income,with possible values >50K or50K,O for50K/50K/50K,0 if
BUSS6002 Individual Assignment Semester 2, 2025 Instructions • Due: at 23:59 on Thursday, 30 October, 2025 (week 12). • You must submit a written report (in PDF, under Canvas-Assignment-Individual Assign- ment (PDF)) with the following filename format, replacing STUDENTID with your own student ID: BUSS6002 STUDENTID. pdf. • You must also submit a Jupyter Notebook file ( . ipynb, under Canvas-Assignment-Individual Assignment (ipynb)) with the following filename format, replacing STUDENTID with your own student ID: BUSS6002 STUDENTID . ipynb. • There is a limit of 6 A4-pages for your report (including equations, tables, captions and reference). • Your report should have an appropriate title (of your own choice). • Do not include a cover page. • All plots, computational tasks, and results must be completed using Python. • Each section of your report must be clearly labelled with a heading. • Do not include any Python code as part of your report. • All figures must be appropriately sized and have readable axis labels and legends (where applicable). • The submitted . ipynb file must contain all the code used in the development of your report. • The submitted . ipynb file must be free of any errors, and the results must be reproducible. • You may submit multiple times but only your last submission will be marked. • A late penalty applies if you submit your assignment late without a successful special con- sideration (or simple extension). See the Unit Outline for more details. • Late submission. In accordance with University policy, these penalties apply when written work is submitted after 11:59pm on the due date: — Deduction of 5% of the maximum mark for each calendar day after the due date. — After ten calendar days late, a mark of zero will be awarded. • Generative AI tools (such as ChatGPT) could be used for this assignment but you should add a statement at the end of your report specifying how generative AI was used. E.g., Generative AI was used only used for editing the final report text. See UoS outline section “Use of generative artificial intelligence (AI)” for detailed instructions. Description The Chicago Board Options Exchange (CBOE) Gold Exchange-Traded Fund (ETF) Volatility Index (GVZ) measures the markets expectations of near-term volatility in gold prices, derived from option prices on the Standard & Poors Depositary Receipts Gold Shares ETF. Predicting the GVZ is useful because it helps investors anticipate fluctuations in gold market volatility and manage portfolio risk more effectively. In this assignment, you are conducting a study that compares the predictive performance of four families of basis functions: piece-wise constant, piece-wise linear, radial, and Laplace, within a linear basis function (LBF) model designed to predict the GVZ index value. The objective is to determine which family of basis functions is most suitable for modelling the relationship between time and gold market volatility (measured by GVZ). You are provided with the GVZ dataset, sourced from the Federal Reserve Economic Data, Federal Reserve Bank of St. Louis. The dataset contains daily observations of GVZ values (GVZ) from 2008 to 2025, along with the Year-Month and Month Index for which the values are recorded. You will be working with the Month Index (as the independent variable x in regression) and the GVZ (as the dependent/response variable y in regression). The actual (Year-Month) is also provided so that you can match the month index value with the actual month of related event in the history, such as the 2008 Global Financial Crisis, to facilitate your understanding of the economic implications of the GVZ index. A scatter plot of the dataset is shown in Figure 1. Figure 1: GVZ levels from June 2008 to Sep 2025. The specific LBF model being considered in your study is given by y = φ(x)Tβ + ε, where y is the GVZ index value, x is the month index, and ε is a random noise; φ(x) denotes the vector of basis function values; the parameter vector to be estimated is β . Four families of basis functions are considered for computing φ(x). Piece-wise constant basis function The first family is the set of piece-wise constant basis functions φ(x) := [1,γ1 (x),...,γk (x)]T , with γi (x) := I(x > ti ), where I(x > ti ) is an indicator function defined by The break points are calculated according to (1) where xmin and xmax denote the smallest and largest observed values of x, respectively. Piece-wise linear basis function The second family is the set of piece-wise linear basis functions φ(x) := [1, x,λ1 (x),...,λk (x)]T , with λi (x) := (x - ti )I(x > ti ), where ti is given by Equation (1). Radial basis function The third family is the set of radial basis functions φ(x) := [1,ρ1 (x),...,ρk (x)]T , with where ti is given by Equation (1). Laplace basis function The final family is the set of Laplace basis functions φ(x) := [1,τ1 (x),...,τk (x)]T , with where ti is given by Equation (1). Before comparing the four basis function families, you must set the number of components k for all models. This hyperparameter value for each basis function family should be selected using a validation set, by minimising the validation mean squared error (MSE). You should select the optimal values of k by exhaustively searching through an equally-spaced grid from 1 to 30, with a spacing of 1: K := {1, 2, . . . , 30}. Once the optimal values of the hyperparameters are chosen for all basis function families, you will be able to compare the predictive performance between the four using a test set (i.e., by comparing the test MSE between the four optimally selected models). With respect to the train-validation-test split, you should use the data points with month index 1-150 as the train set; 151-180 as the validation set; and 181-208 as the test set. Report Structure Your report must contain the following four sections. The number of pages for each section is indicative only and not compulsory. Report Title 1 Introduction (approximately 0.5 pages) — Provide a brief project background so that the reader of your report can understand the general problem that you are solving. — Motivate your research question. — State the aim of your project. — Provide a short summary of each of the rest of the sections in your report (e.g., “The report proceeds as follows: Section 2 presents . . . . Section 3 shows”). 2 Methodology (approximately 2.5 pages) — Define and describe the LBF model. — Define and describe the four choices of basis function families being investigated. — Describe how the parameter vector β is estimated given the value of the hyperparameter k. Discuss any potential numerical issues associated with the estimation procedure. — Describe how the hyperparameter value can be determined automatically from data (as opposed to manually setting the hyperparameter to an arbitrary value). — Describe how the performance of the four families of basis functions is compared given the optimal hyperparameter value. 3 Empirical Study (approximately 2.5 pages) — Describe the datasets used in your study and discuss your observations for the data. — Present (in a table) the selected hyperparameter value for each basis function family. — Describe and discuss the table of selected hyperparameters. — Visually present (using plots) the predicted response values for each basis function family in the test set. — Describe and discuss the plots of predicted values. — Present (in a table) the test MSE values for each basis function family. — Describe and discuss the table of test MSE values. — Report the GVZ forecasts of Oct 2025, Nov 2025, Dec 2025, given by the model with the smallest test MSE. Include a brief description of how these forecasts are obtained. After completing the unit, at the end of 2025 you could compare your forecasts to the actual observations and see how accurate they are. 4 Conclusion (approximately 0.5 pages) — Discuss your overall findings / insights. — Discuss any limitations of your study. — Suggest potential directions of extending your study. Marking Rubric This assignment is worth 30% of the unit’s marks. The assessment is designed to test your compu- tational skills in implementing algorithms and conducting empirical experiments, as well as your communication skills in writing a concise and coherent report presenting your approach and results. The mark allocation across assessment items is given in Table 1. Table 1: Assessment Items and Mark Allocation
ELEN90066 Embedded System Design Kobuki Obstacle Course Project Report November, 2019 1 Introduction In this Kobuki obstacle course project report, we document, explore and discuss all the relevant compo- nents which are part of this project. Particularly, we will discuss our Finite State Machine (FSM), sensor data usage, design, testing, and validation procedures. After that, we will discuss our final workshop outcome and some of the unexpected events which took place. This will be followed by our reflection and ideas for future improvements. The Kobuki robot is best described as a low-cost mobile research based machine which is tailored for education and research on robotics. It has broad range of basic on board sensors and actuators. The factory calibrated gyroscope enables it to perform. precise and accurate navigation. In the literature, the Kobuki robot has been used for analysing cooperative localisation [12], investi- gating the design of low-cost autonomous robot platform. [10], performing people detection and tracking with mounted Microsoft Kinect [2], and also case study for robot modelling and control [11]. It is clear that the Kobuki is a popular option among educators and researchers as it is flexible, capable and more affordable than other comparable robotics systems. While we are not mounting any additional or more complex sensors on the Kobuki robot itself, its on-board sensors should be sufficient for our task of navigating the obstacle course. 1.1 Aim and Objectives The aim of the project is to design and implement a FSM that is capable of tackling the designated Kobuki obstacle course as outlined in the project brief. Other objectives are listed as follows: 1. To learn, apply, iterate upon the concepts and knowledge gained from the Embedded System Design course. 2. To improve our familiarity with programming the Kobuki robot through C (Eclipse IDE) and Lab- view programming environments. 3. To test and validate our design using the available infrastructure and software. 4. To document and describe the processes we went through and recommend potential future im- provements. 2 FSM Diagram The style and notations ofthe FSM attempt follow the ones shown in the course lectures and the assigned textbook [6] where possible and applicable. Please note that after each state transition, the startAngle and startDist variables are updated to netAngle and netDist values. So that we can keep track of the relative angle and relative distance changes. However, this is not explicitly stated in the FSM to help keep the diagram more readable. Figure 1: Finite State Machine In addition, please note that Direction is a Boolean variable that is either Left (0) or Right (1). Ini- tially, if the LeftBump, CentreBump, LeftCliff, CentreCliff is activated, then the direction is set to Right. If the RightBump or RightCliff is activated, then the direction is set the Left. 3 Sensor Data Usage Kobuki robot provides a wide range of basic on-board sensors which can be accessed and processed by the myRIO. Figure 2 provides a an overview of some of the sensors available. However, since not all the sensors are directly applicable or useful for tackling the obstacle course, therefore only a subset of all sensors are used for design and implementation. The sensors used from Kobuki are listed as follows. • Bump Sensor (Left/Centre/Right) This is provided as a Boolean data type, and when any of the bump sensor is triggered, the vari- able would go from false to true and activate the obstacle avoidance mode. The update is then pro- cessed by the guard of the FSM so that the robot would know which direction to start searching. For instance, if the left or centre bumper is hit, then the robot will start performing an exhaustive search to the right hand side. On the other hand, if the right bumper is hit, then the robot will perform an exhaustive search to the left hand side. • Cliff sensor (Left/Centre/Right) Similar to the bump sensor, the cliff sensor is provided as a Boolean data type. When a cliff is detected, the sensor data updates from false to true and the robot enters obstacle avoidance mode in the FSM. The same logic of deciding which side to search is used as in the bump sensor scenario. • Wheel Drop Sensors (Left/Right) The sensor data here is provided as Boolean data type. When the left wheel drop sensor is trig- gered, the robot enters obstacle avoidance mode in the FSM, and begins search to the right. Con- versely, when the right wheel drop sensor is triggered, the robot will search to the left. • Gyroscope The sensor data for gyroscope is provided as an integer in the form of 100deg. This means that the gyroscope data needs to be divided by 100 first to convert it into degrees. The gyroscope sensor is mainly used to correct the robot’s heading when it is in the Drive state. The reason behind this is that, if we only relied on the encoders of the robot the whole time (dead reckoning), then errors can accumulate over time from slip. Significant heading error can also arise when the robot makes contact with an obstacle on the side but does not trigger the left or right bump sensors, this could drag the robot side way and alter the overall heading. To address this, a simple proportional con- troller is implemented to track the initial heading direction of 0。, which is initialised and recorded right after the Labview project is deployed. • Encoders While the encoder raw measurements are available in ticks, the Labview statechart project already provides access to the processed variables Inputs.net distance (mm) and Inputs .net angle (deg). This is used to track the amount that the robot has turned and travelled, and is placed across majority of the guards in the FSM. • Accelerometer (myRIO on-board sensor) Although Kobuki robot does not have a built in accelerometer, acceleration data in (g) can still be measured using the accelerometer within the myRIO. The acceleration data is mainly used to determine if the robot is driving on a flat surface, upward incline, or downward slope. For example, the guard we have in place between the Drive Level and Drive Up state is jacc.xj + jacc.yj > 0.2 (1) which allows the robot to determine that it is driving up an incline when the sum of the abso- lute acceleration values in x and y axes is greater than a fixed threshold. The fixed threshold is determined experimentally during the workshop sessions. Other guards in the FSM utilising the acceleration measurements operate on the same principle, but just with different equality signs and thresholds. While we initially attempted to just utilise the z-axis acceleration data to determine the presence of a slope, but the change in value between flat surface and the slopes we tested was too small. Hence, it was difficult to rely purely on z-axis measurement for robust performance. Please note that sensor data acquisition tasks mainly are based off the tutorials from the course assigned lab manual [4], and also with reference to the official Kobuki online documentation [5]. Figure 2: Kobuki robot with labelled sensors 4 Design Procedure 4.1 Signal Filtering In the first workshop, we learnt about establishing Bluetooth connection to a Nintendo WiiMote and how to send commands to stream the corresponding accelerometer data. More importantly, we also learnt about how to apply a low pass filter to filter out the high frequency noise of the accelerometer sensor. The filter introduced was a relatively simple low pass filter that is introduced in the lab manual [4], which links to another document regarding implementation of tilt-compensated e-compass using accelerometer and magnetometer sensors [9]. While the other parts of the lab was not really re-used for later parts of our design process, this basic filter turned out to be surprisingly useful throughout the course. This is mainly due to its simplicity and ease of implementation. The different equation for filtering the input series x[n] into output y[n] is given by equation 2. y[n] = (1 - )y[n - 1] + x[n] (2) For the case where this low pass filter is applied to accelerometer measurement, y[n - 1] would be the previous filtered value, x[n] would be the current unfiltered input, and y[n] would be the filtered acceleration measurement. Variable is a tunable parameter between 0 to 1 which is adjusted based on experimental results. • If is large (closer to 1), then there will be minimal smoothing effect, as a significant weighting is given to the current iteration input. • If is small (closer to 0), then there will be a large smoothing effect, as a significant weighting is given to the previous filtered value. This filter is later applied to the myRIO accelerometer output measurements to filter out the high fre- quency noise. The parameter = 0.3 was chosen for this purpose. 4.2 Basic Obstacle Avoidance In workshop 4, we started implementing a basic obstacle avoidance algorithm for navigating around a simple block of obstacle. The main idea we had for this was, depending on which bumper was hit first, the robot will choose an initial direction and then attempt to navigate around the obstacle. • If the left or centre bumper is hit, the initial direction will be set to right. • If the right bumper is hit, the initial direction will be set to left. The steps that the robot follow is that, after it has bumped into an obstacle, it will immediately reverse for a set distance, then turn 90。right (or left, depending on which bumper is hit), go forward a set dis- tance, then turn left (or right), go forward and continue its navigation. This forms the basis for our FSM design as we continue to develop and build on top of this idea in the following workshops. If the Kobuki is still unable to move forward, it will follow the same turning procedure, continuously turning right (or left) until it has either: • Found a gap to move through. • Cannot move in that direction anymore, hence it will follow the same steps except search for an opening in the opposite direction. If no such opening is found, the Kobuki will turn back from where it came on. 4.3 Hill Climb Implementation In workshop 5, we were introduced various methods determining whether the Kobuki robot is situated on an inclined surface. The application note provided by Analog Devices [1] listed a breadth of options for how the tilt angle could be determined using an accelerometer, which includes: • Single-Axis Tilt Calculation AX[g] = 1g × sin(θ) (3) where X is the horizontal axis and we can arrange the equation below and solve for the inclination angle θ = arcsin(AX). • Dual-Axis Tilt Calculation (4) where X is the horizontal axis and Y is the vertical axis. • Three-Axis Tilt Calculation (5) where X and Y are the horizontal axes and Z is the vertical axis. During the workshop testing phase, we first implemented the single axis tilt calculation method, and found that the while this approach was simple and easy to implement, it was not very robust and sometimes would not trigger the guard which we had in place for incline detection. After that, we tried to implement our own hybrid approach between the dual and three-axis tilt calculation. Our incline detection method to directly compute an angle, but to use a combination of acceleration measurement and a fixed threshold to perform incline detection. This relies on the abso- lute values of the acceleration in x and y axes, which are denoted as acc .x and acc .y. As previously mentioned, the guard between Drive Level and Drive Up (for an incline), is as follows. jacc.xj + jacc.yj > 0.2 (6) Assuming that the x-axis is pointing up the incline, and y-axis is point to the side of the incline, as shown in figure 3. Then regardless of which orientation Kobuki approaches the incline, the myRIO accelerometer will be able to compute the sum above and determine if an incline exists. Figure 3: Incline • When the robot is on flat surface, the contribution from acc .x and acc .y are both 0. • When the robot is on an incline as shown in figure 3, then the accelerometer value in acc.x would increase, while acc .y remains approximately 0. • When the robot is on an incline, but oriented at an angle, then acc .x and acc .y would both have non-zero values. 4.4 Choosing Development Environment We decided to use the statechart module on LabView over developing the FSM in C as the statechart interface was more intuitive when looking at how the FSM transitions from one state to another. The myRIO is also built around the use of LabView, making it easier to upload and deploy code onto the hardware itself. The team’s experience with LabView over the C development IDE also played a factor into this decision. 4.5 Robot Behaviour Refinement One of the requirements for the project if for the Kobuki to maintain ground orientation. The Kobuki may not always keep ground orientation, especially after turning, due to slip or inaccurate encoder read- ings. Hence we have implemented a Proportional Control in the drive state that corrects the Kobuki’s orientation using the error between the gyroscope’s reading and the reference orientation. Below are the equations used to adjust the motor speeds for when the Kobuki is on ground level. Left Motor Speed = 200(1 + P θ(netAngle - AngleRef)) Right Motor Speed = 200(1 - P θ(netAngle - AngleRef)) where P θ = 0.01 is the Proportional gain for the ground orientation. Another requirement is for the Kobuki to drive up the ramp without falling off the edge. We could have let the Kobuki run through the Obstacle Avoidance State every time it reaches the edge of a ramp, but it is not an ideal method for that state to be always triggering, especially if one of the cliff sensors is not working properly. Our refinement to this is to add another Proportional Control that uses the accelerometer readings to have the Kobuki re-orientate itself so that it travels straight uphill. Below shows the equations used to set the motor speeds for when the Kobuki is driving down the hill. Left Motor Speed = 200(1 + Pa(acc.y)) Right Motor Speed = 200(1 - Pa(acc.y)) where Pa = 3 is the Proportional gain for the accelerometer reading. (a) Drive state controller implementation (b) Drive down state controller implementation 4.6 Exhaustive Search Implementation The Obstacle Avoidance state has been fitted with an exhaustive search algorithm. The exhaustive search algorithm simply makes the Kobuki look for all possible openings if its path is blocked. It searches for a path in one direction, then switches directions if it cannot go in tat direction anymore. If no such opening is found, the Kobuki will simply turn back to where it came from. Such direction changes and choice of turning back are made into simple and general diagrams with the help of variables and triggers. Below shows a representation on how the use of variables and triggers help describe the state of the Kobuki. 5 Testing Procedure 5.1 Testing Using CyberSim Simulator From the project handout, a relatively detailed specifications were provided for the course, but we were also told that we would not be able to perform trials on the actual course before the final workshop. Therefore, we started exploring other testing options which are more extensive and complete. The main one which we decided to explore further is the CyberSim simulator which we already have access to. However, CyberSim was not particularly well documented and the existing simulation maps have their configurations stored in .xml files which are difficult to visualise and customise. Upon further research, we realised that all the existing simulation maps can be created and edited using LabVIEW Robotics Environment Simulator. Essentially, LabVIEW Robotics Project Template has configuration wizard that provides a simple and basic interface that allows users to open and edit existing .xml files, from there we were able to modify and add to existing map elements such as walls, obstacles and so on. The main prerequisites for making the simulation map are: • NI LabVIEW 2018 myRIO Software Bundle • CAD software capable of exporting any of the following file types (.ive, .dae, .lvsg, .wrl). It turns out that Fusion360 does not work well for this purpose, and an alternative called Rhino 6 was used for testing and importing custom objects. • Detailed write-up on how this procedure is conducted is available in the appendix section 10.1. From the provided sample obstacle course layout shown in figure 6, we were able to construct a sim- ulation map after making a few assumptions regarding measurements which were not explicitly stated. While it was possible to import CAD models, we were unable to obtain a matching physical/collision model after import. These limitations will be discussed in the following subsection. Therefore, we ended up using the existing components from the Built-in Obstacle Library to construct our simulation map. This consists of sphere, cylinder and box. Figure 8: Simulation Scene (Perspective View) The resulting simulation environment with slight variation in obstacle arrangement can be seen in figures 7 and 8. The main advantage for this approach is that all of the obstacles and walls can be customised by specifying its centre coordinates, dimensions and object properties such as colour and whether it can be moved. One selected test for obstacle avoidance robustness test was recorded [7], and it shows that the robot is capable of performing the exhaustive search algorithm and avoid a wide rectangular block. The main advantage of testing by simulation in CyberSim is that, you can iterate very quickly and do not need to have access to the physical myRIO and Kobuki hardware. 5.2 Limitations of CyberSim Simulation While the aforementioned simulation map design and setup is convenient and relatively easy to modify and iterate, there is also quite a few limitations which are stated as follows. • The inability to import CAD model while matching the physical profile exactly. While the CAD model importer allows us to import models from the specified formats, we were unable to figure out how to apply the physical model so that it would fit the imported shape exactly. For example, if we import a model of a triangular prism, we would need to select a physical model of a box, cylinder or sphere to approximate the collision model. • Imperfect modelling of the robot, leading to unrealistic simulation behaviour. The example that best illustrates this would be the ramp/hill implementation in the simulation environment. In real life, when a Kobuki robot falls or is raised, both of its main wheels have mechanical mechanism that allows the wheels to be dropped. However, this behaviour is not included in the Kobuki model used in CyberSim. Consequently, the simulated Kobuki would struggle to transition from flat surface onto a ramp or incline with more than 5。angle. • The axes of the simulated Kobuki do not align with the the actual Kobuki. Hence, we would need to ensure that the axes used in simulation are updated accordingly before testing in real life. • There is significantly more sensor noise in real life. Therefore, algorithms which work well in CyberSim simulation do not always carry through directly to real life. A potential future improve- ment is to incorporate simulated sensor noise, with tune-able parameters for noise distribution. 5.3 Testing Using Available Infrastructures With the limitations stated in previous subsection in mind, we look to the available physical testing components to help improve the overall robustness of our design. In workshops 7 to 9, we were given the opportunity to improve and refine our FSM design and the actual implementation with the Kobuki robot. The testing procedure consists of the following steps: 1. Update the FSM implementation in Labview statechart, and upload the code to the myRIO via wired or wireless connection. 2. Setup the obstacle course with the available obstacles and ramp. This includes tweaking object orientations, initial robot position and setting up different test cases, so that we can explore as many untested cases as possible. 3. Run the robot and observe its behaviour to see whether the actual behaviour matches the expected behaviour. 4. Repeat steps 1 to 3 to continue the design iterations.
MATH-UA 121: Calculus I Final examination Multiple Choice Shade your answers to the multiple choice questions on the multiple choice answer sheet on page 2. Only multiple choice answers on the multiple choice answer sheet will be graded. No explanation is required to be shown and no partial credit will be given for the multiple choice questions. 1. (2 points) Which of the following is the domain of the function A. (−∞,−2) ∪ (3,∞) B. (3,∞) C. (0,∞) D. (−∞, 3) E. (−2, 0) 2. (2 points) Evaluate A. 2 B. −2 C. 0 D. ∞ E. None of the above 3. (2 points) Find the values of a and b for which f is continuous everywhere: A. a = 0, b = 0 B. a = 0, b = 4/3 C. a = 1, b = 3 D. a = 1, b = 4 E. a = 3/4 , b = 1 4. (2 points) Given that the equation x3 + 5 = −x has exactly one real solution, in which of the following intervals does the solution lie? A. (−2,−1) B. (−1, 0) C. (0, 1) D. (1, 2) E. None of the above 5. (2 points) Evaluate A. − 2/1 B. −2 C. 2 D. 2/1 E. Limit does not exist (is possibly ±∞) 6. (2 points) Differentiate 7. (2 points) At which one of the following values of x is the tangent line to the curve y = 3+5x − x 2 parallel to the line x + y = 3? A. 0 B. 1 C. 2 D. 3 E. 4 8. (2 points) Which one of the following is the derivative of cos(x3) + (sin(x))3? A. −sin(x3) + 3(sin(x))2 B. −3x 2 sin(x3) + 3(sin(x))2 cos(x). C. 3x 2 cos(x3) − 3 sin(x)(cos(x))2 D. cos(x3) + 3(sin(x))2 E. None of the above 9. (2 points) Suppose f is a differentiable function for all real numbers. Suppose f'(3) = −2 and f (3) = 10. Using linearization or differentials, approximate f (2.5). A. 8 B. 9 C. 9.5 D. 11 E. 12 10. (2 points) Evaluate the limit A. 0 B. 1 C. /e D. ∞ E. −∞ 11. (2 points) The horizontal and vertical asymptotes of the function are A. y = 1 and x = 0 B. y = 0 and x = 0 C. y = 1, y = 0, and x = 0 D. y = −1, y = 0, and x = 0 E. None of the above 12. (2 points) Suppose f is a one-to-one differentiable function with values as shown. What is(f−1)'(−2)? A. −2 B. −1 C. − 2/1 D. − 3/1 E. − 4/1 13. (2 points) Find an equation for the tangent line to the ellipse x2 + 2xy + 4y2 = 12 at (2, 1). A. 2y = 4 − x B. 2y = x C. 2y = −x D. y = x − 1 E. y = −2x + 5 14. (2 points) Evaluate f'(3) if f (x) = x 1−x. A. − 9/ln3 − 27/2 B. − 2/ln2 − 4/1 C. 1 D. ln2 + 1 E. ln3 − 1 15. (2 points) What is the derivative of y = arctan(eu)? A. 1 + eu/1 B. 1 + e −2u/1 C. 1 + e2u/eu D. 1 + e −2u/−e−u E. None of the above 16. (2 points) Find the absolute minimum value of on the interval [1, 3]. A. 0 B. 5/2 C. 13/6 D. 2/1 E. 5/4 17. (2 points) The graph below is f'(x) for some function f . (Note: This is the graph of the derivative of f .) Which of the following are true? I. f has a horizontal tangent line when x = 5. II. f is increasing on (−∞,−2) and (4, 6). III. f is decreasing on (−∞, 0) and (5,∞). A. I only B. II only C. III only D. I and II E. II and III 18. (2 points) What is the value of A. 6/1 B. 1 C. 3/8 D. 17/6 E. None of the above 19. (2 points) If and then what is A. 8 B. 10 C. 14 D. 16 E. 17 20. (2 points) True or False: Suppose f is defined at a and limx→a g(x) exists. Then one always has A. True B. False Free Response Please show and explain your working to receive credit. 21. The Mall of America is setting up a Christmas Village as part of its holiday display. It is to be laid out in a rectangular shape with three 6 foot-wide openings. The rest of the border should use 82 feet of fencing. (Diagram not to scale.) (a) (3 points) Let the length and the width of the rectangle be x and y, respectively. Using the constraint on the length of the fence, write y as a function of x. (b) (3 points) Write the area function, A(x), of the holiday display as a function of x. (c) (3 points) Find x that maximizes A(x). (Make sure to justify why your answer maximizes A(x).) (d) (1 point) Find the maximum area. 22. Suppose (a) (2 points) Find g'(x). (b) (2 points) Is g(x) one-to-one? Justify your reasoning. (c) (6 points) Find the intervals where g(x) is concave up and concave down and identify the x values for any inflection points. 23. (10 points) Estimate the area under the graph of f (x) = x3 + 1 from x = 0 to x = 4 using four approximating rectangles and right endpoints. Sketch the graph and the rectangles. Is your estimate an underestimate or overestimate? 24. A particle is moving along a line. It starts off at time t = 0 at position s(0) = 2 and has velocity function v(t) = t2 − t. (a) (4 points) Find the position of the particle at time t = 4. (b) (6 points) Find the distance traveled by the particle during the time interval [0, 4].
