cmput 412 experimental mobile robotics csaba szepesvári university of alberta
TRANSCRIPT
CMPUT 412CMPUT 412Experimental Mobile Experimental Mobile
RoboticsRobotics
Csaba SzepesvCsaba Szepesváriári
University of AlbertaUniversity of Alberta
Plan for todayPlan for today
Topic of the course Introduction/admin Expectations Requirements/Marking Course contents
Topic: Autonomous DrivingTopic: Autonomous Driving
Goal: Cars should drive themselvesGoal: Cars should drive themselves Advantages: Less accidents, increased Advantages: Less accidents, increased
efficiencyefficiency History: History:
1961: Stanford cart 1987-1995: Dickmanns
180km/h, 1000km,human intervention,driving on highways
2005: DARPAGrand Challenge,211 km desert course
20072007
2007: DARPAUrban Challenge
Autonomous parking systems(Lexus, Mercedes, Toyota,..)
Cybercars project
Course projectCourse project
The tasksThe tasks Following prespecified routes in a "city"Following prespecified routes in a "city"
[14 January - 28 January][14 January - 28 January] Task #1: Build a robot that follows a white tape, taped to the floorTask #1: Build a robot that follows a white tape, taped to the floor Task #2: City-like environment, follow a prespecified routeTask #2: City-like environment, follow a prespecified route
Taxiing on demand,Taxiing on demand, obstructions on the roadobstructions on the road [29 January - 25 February][29 January - 25 February] Task #1: Picking up passengers at various locations and taxiing Task #1: Picking up passengers at various locations and taxiing
them to other locations. Uses bluetooththem to other locations. Uses bluetooth Task #2: Parts of the route can become blocked. Task #2: Parts of the route can become blocked.
Parking, dealing with trafficParking, dealing with traffic [26 February - 7 April][26 February - 7 April] Task #1: Parallel parking Task #1: Parallel parking Task #2: Multiple robots on the road at the same time, avoiding Task #2: Multiple robots on the road at the same time, avoiding
collisionscollisions
AdminAdmin
Me..Me..
Studies:Studies: Mathematics (Stat. and Prob.)Mathematics (Stat. and Prob.) Computer ScienceComputer Science
ResearchResearch Reinforcement learning (theory)Reinforcement learning (theory) Machine learning (vision, robotics,..)Machine learning (vision, robotics,..)
ExperienceExperience 5 years in industry (sw firm, 5 years in industry (sw firm, speech, text, videospeech, text, video)) 15 years of C++15 years of C++
You..?You..?
Fill out form!Fill out form!
ScheduleSchedule!!
Lecture: TR 15:30-17:00Lecture: TR 15:30-17:00 Room: ETL E1 018Room: ETL E1 018
Lab: W 14:00-17:00Lab: W 14:00-17:00 Room: CSC 229 Room: CSC 229
Office hoursOffice hours
Appointment:Appointment: [email protected]@cs.ualberta.ca
Stop by!Stop by! Room: Ath 3-11Room: Ath 3-11 Phone: x2-8581Phone: x2-8581
Information sourcesInformation sources
Course webpage:Course webpage:ugweb.cs.ualberta.ca/~c412ugweb.cs.ualberta.ca/~c412
RLAI page:RLAI page:http://rlai.cs.ualberta.ca/openpages2/CMPhttp://rlai.cs.ualberta.ca/openpages2/CMPUT412+2008UT412+2008
Teaching Assistants: Teaching Assistants: Azad Shademan, Neesha DesaiAzad Shademan, Neesha Desai
Instructional group support:Instructional group support: John Rodson (John Rodson ([email protected]@ugrad.cs.ualberta.ca))
PrerequisitesPrerequisites
No official prerequisitesNo official prerequisites Brush up your knowledge in:Brush up your knowledge in:
Calculus (MATH 114,115, 214)Calculus (MATH 114,115, 214) Linear algebra (MATH 120 or 125)Linear algebra (MATH 120 or 125) Probability & stat (STAT 221)Probability & stat (STAT 221)
Good to know aboutGood to know about Machine learning (CMPUT 466/551)Machine learning (CMPUT 466/551)
Programming:Programming: C/C++C/C++
ExpectationsExpectations
GoalsGoals
Learn about roboticsLearn about robotics ConceptsConcepts TechniquesTechniques ChallengesChallenges
Complete the projectComplete the project Have fun!Have fun!
Course formatCourse format
Formal lectures (~3 weeks)Formal lectures (~3 weeks) BackgroundBackground
Weekly project meetingsWeekly project meetings Discussion of progressDiscussion of progress ChallengesChallenges Planning (to meat deadline)Planning (to meat deadline)
Steady workloadSteady workload
My contributionMy contribution
I act like a supervisorI act like a supervisor I define tasksI define tasks I provide background material (lectures, I provide background material (lectures,
resources)resources) I answer questionsI answer questions I evaluate your performanceI evaluate your performance
My expectationsMy expectations
You You try your besttry your best to solve the assignments (it’s to solve the assignments (it’s fun!)fun!)
You will act in a You will act in a self-initiatedself-initiated manner manner You ask You ask questionsquestions Cooperate, but contributeCooperate, but contribute You come to the lecturesYou come to the lectures You come to the labsYou come to the labs NoNo cheating, plagiarism, misrepresentation of cheating, plagiarism, misrepresentation of
facts (see course webpage for detailed info)facts (see course webpage for detailed info)
MarkingMarking
How to get (good) marks?How to get (good) marks?
This is a This is a
project based courseproject based course=> no final, or midterm=> no final, or midterm
““Just” solve the problemsJust” solve the problems
MarkingMarking
No predefined grading systemNo predefined grading system Individual performances: Based on the reports, Individual performances: Based on the reports,
participation in the meetings, presentations, lecturesparticipation in the meetings, presentations, lectures You are going to evaluate each others’ presentationsYou are going to evaluate each others’ presentations
Course Work Due Report weight
Presentation weight
Assignment 1 J an 28 20% 10%
Assignment 2 Feb 25 20% 10%
Assignment 3 April 7 30% 10%
(Your) class presentations(Your) class presentations
Be preparedBe prepared Keep structure:Keep structure:
Problem definitionProblem definition Proposed solutionProposed solution EvaluationEvaluation ConclusionsConclusions
Use slidesUse slides Keep time limits [20 min]Keep time limits [20 min]
ReportsReports Title, authorsTitle, authors Introduction (max. 1 page): Introduction (max. 1 page):
the task and the challenges facedthe task and the challenges faced Proposed solution (max. 1 page + figs): Proposed solution (max. 1 page + figs):
possible alternatives, the decisions you took, how you arrived at possible alternatives, the decisions you took, how you arrived at them, measurements, data, related workthem, measurements, data, related work
Evaluation (max. 1-2 pages + figs): Evaluation (max. 1-2 pages + figs): Evaluate your solution. Show to what extent it achieves the goal, Evaluate your solution. Show to what extent it achieves the goal,
describe its limitationsdescribe its limitations Conclusions (max. 1 page) Conclusions (max. 1 page)
summarize your worksummarize your work Sharing the work (1 page):Sharing the work (1 page):
Who did what, how was the time spentWho did what, how was the time spent Format: Format:
Preferably use LaTeX, high quality, compressed presentationPreferably use LaTeX, high quality, compressed presentation
0th assignment0th assignment
Task #1: SUBSCRIBE to the course’s Task #1: SUBSCRIBE to the course’s open pages!open pages!
Task #2: Learn to use the open pagesTask #2: Learn to use the open pages
Open pages:Open pages: Anyone can edit them (in the browser)Anyone can edit them (in the browser) Send notification to subscribersSend notification to subscribers Add comments/questions to the end of the pageAdd comments/questions to the end of the page