1. CS 326A: Motion Planning Jean-Claude Latombe CA: Aditya Mandayam

2. Motion planning is the ability for an agent to compute its own motions in order to achieve certain goals. All autonomous robots and digital actors should eventually have this ability

3. Piano Mover’s Problem

8. PlanMoveSense

9. ARL Robot Goal

11. PlanMoveSense Learn Motion library

12. Goal of Motion Planning Compute motion strategies, e.g.: –geometric paths –time-parameterized trajectories –sequence of sensor-based motion commands To achieve high-level goals, e.g.: –go to A without colliding with obstacles –assemble product P –build map of environment E –find object O

13. Fundamental Question Are two given points connected by a path? Valid region Forbidden region

14. Fundamental Question Are two given points connected by a path? Valid region Forbidden region E.g.: ▪Collision with obstacle ▪Lack of visibility of an object ▪Lack of stability

15. Basic Problem  Statement: Compute a collision-free path for a rigid or articulated object among static obstacles  Inputs: Geometry of moving object and obstacles Kinematics of moving object (degrees of freedom) Initial and goal configurations (placements)  Output: Continuous sequence of collision- free robot configurations connecting the initial and goal configurations

16. Is It Easy?

17. Tool: Configuration Space Problems: Geometric complexity Space dimensionality

18. Continuous space Discretization Search C-space Sampling-basedCriticality-based

19. Extensions of Basic Problem Moving obstacles Multiple robots Movable objects Assembly planning Goal is to acquire information by sensing –Model building –Object finding/tracking –Inspection Nonholonomic constraints Dynamic constraints Stability constraints Optimal planning Uncertainty in model, control and sensing Exploiting task mechanics (sensorless motions, under- actualted systems) Physical models and deformable objects Integration of planning and control Integration with higher- level planning

20. Some Applications

21. Humanoid Robots HRP-2, AIST, Japan

22. Lunar Vehicle (ATHLETE, NASA/JPL)

23. Climbing Robot http://www.youtube.com/watch?v=biSx-aKN690

24. Dexterous Manipulation

25. Modular Reconfigurable Robots

27. Manipulation of Deformable Objects Topologically defined goal

28. Digital Characters A Bug’s Life (Pixar/Disney) Toy Story (Pixar/Disney) Tomb Raider 3 (Eidos Interactive)Final Fantasy VIII (SquareOne)The Legend of Zelda (Nintendo) Antz (Dreamworks)

29. Digital Characters

30. Animation of Crowds

32. Design for Manufacturing and Servicing

35. Assembly Sequence Planning

37. Cable Harness/ Pipe design

38. Map Building Where to move next?

39. Navigation Through Virtual Environments

40. Virtual Angiography / Bronchoscopy / Colonoscopy

41. Radiosurgical Planning CyberKnife (Accuray)

42. Building Code Verification 9-inch turning radius24-inch turning radius

43. Egress Simulation Primary escape route Secondary escape route Potential congesting areas

44. Self-Parking

45. Transportation of A380 Fuselage through Small Villages Kineo

46. Study of Motion of Bio-Molecules Inhibitor binding to HIV protease

47. Goals of CS326A  Present a coherent framework for motion planning problems  Emphasis of “practical” algorithms with some guarantees of performance over “theoretical” or purely “heuristic” algorithms

48. General Framework Continuous representation (configuration space and related spaces + constraints) Discretization (probabilistic sampling, criticality-based decomposition) Graph searching (blind, best-first, A*)

49. Practical Algorithms (1/2)  A complete motion planner always returns a solution plan when one exists and indicates that no such plan exists otherwise.  Most motion planning problems are hard, meaning that complete planners take exponential time in # of degrees of freedom, objects, etc.

50. Practical Algorithms (2/2)  Theoretical algorithms strive for completeness and minimal worst-case complexity. Difficult to implement and not robust.  Heuristic algorithms strive for efficiency in commonly encountered situations. Usually no performance guarantee.  Weaker completeness  Simplifying assumptions  Exponential algorithms that work in practice

51. Prerequisites for CS326A  Ability and willingness to complete a significant programming project with graphic interface.  Basic knowledge and taste for geometry and algorithms.  Interest in devoting reasonable time each week in reading papers.

52. CS326A is not a course in … Differential Geometry and Topology Kinematics and Dynamics Geometric Modeling … but it makes use of knowledge from all these areas

53. Work to Do A.Attend every class B.Prepare/give two presentations with ppt slides (20 minutes each) C.For each class read the two papers listed as “required reading” in advance D.Complete the programming project E.Complete two homework assignments

54. Programming Project Navigate in virtual environment Simulate legged robot Inspection of structures Search and escape

55. Website and Schedule ai.stanford.edu/~latombe/cs326/2009/index.htm