2/1/16 My Research CMSC 691 Spring 2016 2 Artificia Ar ificial l in intellig elligen ence ce How w to o get et comp computer ers s to o beh ehave e “in “intellig elligen ently”? y”? Rob obot otics ics Na Natural l Language e Processin rocessing Sen ensin sing the e en environ ironmen ment, How w ca can comp computer ers s under erst stand huma man la languages es (E (English lish, Hin indi, i, doin oing physica ysical l op oper eration ions, s, et etc.), ), rea read, sp spea eak and list listen en, build ildin ing hard rdwa ware re, pla lannin ing Principles of motion mot ions, s, perf erfor ormin ming task sks perf erfor orm m la language e task sks Human-Robot Interaction Huma man-R -Rob obot ot In Inter eract ction ion an introduction to robots in our daily lives How w ca can we e go o from rom in indust stria ial l rob robot ots s to o usef seful l rob robot ots s in in huma man en environ ironmen ments? s? (S (Sch chools, ools, ca cars, s, homes…) omes…) Dr. Cynthia Matuszek Today’s Class What is a Robot? 3 4 u What is HRI? u Autonomous? u Humanoid? u Mobile? u Or a robot, for that matter u Physical? u Sensory? u Manipulative? u Introductions u Department? Interests? Why this class? u Human-friendly? u Intelligent? u What else? u Bookkeeping u Academic honesty policy Alternate theory: Robotics is what happens when “A robot is a reprogrammable, multifunctional u My expectations computers become physical agents – gain the ability manipulator designed to move material, parts, tools, to sense the world (see, hear, etc.), interact physically or specialized devices through variable programmed u Syllabus and schedule, topics we will cover with objects and people, and operate in the physical motions for the performance of a variety of u Class structure and grading environment as well as the virtual world. tasks. ” (Robot Institute of America) u Goals And in Fact In Fiction… 5 6 Wall•E: 2008 ED-209. Robocop: 2014 Sentinel. X-Men, Days of Future Past: 2014 Optimus Prime. Transformers: 2007-current Data. Star Trek: TNG: 1987 1
2/1/16 Human-Robot Interaction Introductions I’m a Honda Asimo! 7 8 u So what do we expect from physical computers? u What do you go by? u In a factory? In a warehouse? On Mars? u Department u In a human-centric environment? u Taking this class because… u The changing newest face of robotics u Research interests u If any u Anything else we should know? u E.g., “I am not a Honda Asimo” Goals of This Course Class Structure 9 10 u Understand HRI u One semest semester er-lon -long project roject u What is the field about? u Thinking, implementing, writing, and presenting the H and the R u [Why] is it important? u Remainder of the course is research paper based u Introduce the huge range of current HRI research u Per topic: u Gain experience conducting HRI research u 2-3 papers u Learn to identify HRI problems in your own research u One person will sign up to presen resent and lea lead discu iscussion ssion u Perform a project involving real human-robot interaction u Ever eryon one will read and discuss u How we will accomplish that u Posting and reading class forums, as specified on to u Paper reading/presentation/discussion bookkeeping... u No exams, occasional small bookkeeping homeworks u Project Expectations Academic Integrity 11 12 u Do o the e rea readin ing u I hate policing students, and I hope not to have to do so. u But, it is extremely unfair to the other students, so… u Participate u I take integrity very seriously. u Participate act ctiv ivel ely in class discussions u Fa Fabrica ication ion: Fabricating sources or any other information in your u Let other people participate, and listen attentively assignments is academically dishonest. u Ask questions! u Aid Aidin ing and abet ettin ing: Providing another student with answers, or helping them to cheat. u Read and post to the forum as specified u Pla lagia iarism ism: Using a source (for code, blocks of text, images, or designs) without appropriate citations and recognition. u Attend class regularly u Cop opyin ying: Using another student's work for an assignment, exam, or project without acknowledgment. u If you will miss class for a good reason (work commitment, religious holiday, serious illness), drop me an email u You can do a LOT of collaboration in this class! 2
2/1/16 Samples of Project Types Project Milestones 13 14 u Evaluation of human factors Research question 1. u In-person or crowd-sourced user studies User study design 2. u Statistical analysis Technical implementation and first demo 3. u Novel HRI system u Robot is doing something new and different User study analysis 4. u The “human” element can be environmental Presentation to professor 5. u Pure research Write-up 6. u Active learning from human interaction Final Presentations 7. MOST of these will occur on the actual robots – may require careful group selection! Grading Policies 15 16 Paper presentation & 25% u This class is primarily paperless leading discussion u Time management Discussion participation 20% u Extensions are sometimes available. Please talk to me! Project 50% u Be courteous to classmates and instructors Research questions 5% • Study design 5% • u No devices in use during presentations • Technical implementation 10% • User study analysis 10% u Whyyyyyy??? • Write-up 10% u You don’t learn as much. • Presentation 5% • Demo & Video 5% u Really. It’s science. I am willing to talk about this, u Peop eople le arou round you ou don on’t lea learn as s much ch. Extra credit (up to) 3% after you read: u But L ? Miscellaneous (surveys etc.) 5% http://tiny.cc/devices-in-class Syllabus Teamwork 17 18 u UP BY MONDAY. u Projects may be done individually or in groups of 2-3 Up tomorrow. u Self-chosen groups u http://tiny.cc/hri-syllabus u I will review u Links to schedule u Teams will: u Coursework, grading, and policies u Meet regularly u Office hours: u Share effort on project elements reasonably u Me: Monday 9:00-10:00pm, Tuesday 10:00-11:00am u Frequently describe who’s doing what u Nisha: Monday 1-2, Wednesday 3
2/1/16 Communications Communications 19 20 u For help: u Developing this class is an ongoing project. u Post to the forum u I genui nuine nely y val alue ue yo your ur cont ntinual nual input nput! u Come to office hours u Suggestions/criticisms/complaints/compliments? u Drop by ITE 331 ß this is actually a great choice Policies Assignments Grading u Drop by ITE 343 ß this too Course Structure Workload Groups u Email: In-class Activities Topics Readings u 24-hour cooling period u How? u Email Dr. M and and Nisha u Post on the forum, email me directly, email/talk to u Email must include a link to your forum post Nisha, slip a note under my door, smoke signals, … HRI Topics Some High-Level Topics 21 22 u HRI taxonomy and metrics u Remote tele-operation u Where will we begin seeing robots? u HRI methods u Expressive robot motion u What is a “Human-centric environment”? Task?” u Assistive robots u Navigation around humans u And how is it relevant to robots? u Socially-assistive robots u Perception of humans u Social robots u Human-robot collaboration u Social learning u How can robots be social beings? When do we want them to? u How can robots express emotion – and (when) should they? u Collaborative manipulation, u Non-verbal human-robot handover communication: Gestures u Human-robot collaboration – how and when? u Natural-language u Human-robot dialog u Assistive robots – helping people with special needs interactions with robots u Remote presence u Robot ethics – what should and shouldn’t robots do? For Next Class What Should They Do? 23 24 u Robots are moving away from factory floors to… Rea ead this is article icle throu rough ch chapter er 5: 5: u Entertainment, toys u Homes (personal robotics) u http://tiny.cc/goodrich-survey u Medical, surgery u Industrial automation (mining, harvesting, warehouses, …) u Hazardous environments (space, underwater, battlefields, …) u I will lead the next discussion u Roads u Read the syllabus u Research Trends u http://tiny.cc/hri-syllabus u Manipulation of everyday objects u Complex household tasks u Read the academic integrity page u Object recognition, mapping, interaction u Human robot interaction 4
2/1/16 5
Recommend
More recommend