Spring 2017 CIS 493, EEC 492, EEC 592: Autonomous Intelligent Robotics Instructor: Shiqi Zhang http://eecs.csuohio.edu/~szhang/teaching/17spring/
Reading ● Be specific on where you are pointing to Good – “In Figure 1-c, it is said that… I don’t understand..” Bad – “It is said in the paper… Why does…? ”
Reading: “BWIBots: A platform for bridging the gap between AI and Human-Robot Interaction research” ● Planning using action language BC – Will be discussed in “task planning” ● Incorporating uncertainty into planning – Will be discussed in “commonsense reasoning” ● Understanding natural language requests – Will be discussed in “natural language processing”
● Robot should be able to notify its admin via email or SMS when it is stuck and can not find a path to its goal ● BWIBot can also be used as telepresence robot – Utkarsh "Virtour: Telepresence System for Remotely Operated Building Tours" https://apps.cs.utexas.edu/apps/sites/default/files/tech_reports/TR-1611.pdf
What type of robot do you need? https://www.ted.com/talks/a_robot_that_flies_like_a_bird At 2:00
Indoor wheeled robots 3k-6k
Indoor wheeled robots 3k-6k 30k-60k
Indoor wheeled robots 3k-6k 30k-60k 300k-600k
Indoor wheeled robots 3k-6k 30k-60k 300k-600k How are they different?
BWIBot hardware
What exactly is happening? Video
Symbolic-level planning: goto(room_101) Semantic map 2d position: <x, y, theta> Global path planner 2d trajectory Local path planner Control signals
Questions ● Which level robot localization is placed at? – Particle filter (AMCL) – Kalman filter ● Where is obstacle avoidance? – Dynamic window – Vector Field Histogram ● Where is A* and/or Dijkstra?
The software architecture for the BWIBots
The next reading Fox, Dieter, Wolfram Burgard, Frank Dellaert, and Sebastian Thrun. ● "Monte carlo localization: Efficient position estimation for mobile robots." AAAI 1999, no. 343-349 (1999): 2-2. This paper just won the AAAI Classic Paper Award ● https://rse-lab.cs.washington.edu/2017-aaai-classic-paper-award-goes-to-monte-carlo-localization-paper/
Recommend
More recommend
