SLIDE 1
Q.-C. Pham
NTU, Singapore
Find a collision-free path between q
s t a r t and q g o a l
Integrating Dynamics into NTU, Singapore Industrial Motion Planning - - PowerPoint PPT Presentation
Integrating Dynamics into NTU, Singapore Industrial Motion Planning - - PowerPoint PPT Presentation
Q.-C. Pham Integrating Dynamics into NTU, Singapore Industrial Motion Planning Path planning problem Find a collision-free path between q s t a r t and q g o a l From academic breakthroughs... Configuration space formulation (Lozano-Perez
SLIDE 2
SLIDE 3
Configuration space formulation (Lozano-Perez 1983) Sampling-based planning (Kavraki et al 1996, Lavalle and Kuffner 2000) Efficient implementations
- ROS / MoveIt !
- OpenRAVE
From academic breakthroughs...
SLIDE 4
... to industrial successes
SLIDE 5
Torque constraints Friction constraints Fluid constraints ZMP constraints
How about dynamics ?
SLIDE 6
Planning in the state space ?
–
More dimensions (2n)
–
Obstacle avoidance difficult to guarantee
–
Less intuitive Trajectory decoupling (path + parameterization)
–
Cluttered environments
–
Can use regular PRM/RRT + many heuristics
–
Optimal time parameterization (Bobrow 1985 and many others)
Planning with dynamics ?
SLIDE 7
Developed by Bobrow (and many others) Applicable to many types of problems
- Velocity / acceleration / torque bounds
- Grip stability / friction constraints
- ZMP constraints
Time-Optimal Path Parameterization (TOPP)
SLIDE 8
Our implementation of Bobrow algorithm
–
https://github.com/quangounet/TOPP
–
Fast (torque constraints 7 DOF, 1s, 100 points : 6ms)
–
Integrated with OpenRAVE
Currently supported constraints
- Velocity / acceleration / torque bounds
- Friction constraints
- ZMP constraints
Time-Optimal Path Parameterization (TOPP)
SLIDE 9
Sampling-based algorithm
SLIDE 10
Sampling-based algorithm
SLIDE 11
Sampling-based algorithm
SLIDE 12
Sampling-based algorithm
SLIDE 13
Sampling-based algorithm
SLIDE 14
Sampling-based algorithm
SLIDE 15
Sampling-based algorithm
SLIDE 16
Sampling-based algorithm
SLIDE 17
Sampling-based algorithm
SLIDE 18
Quasi-static planning
Final path not parameterizable ? Check quasi-static feasibility at each step Loss of completeness / optimality
SLIDE 19
Inputs
- Path in configuration space
- (vmin,vmax) at the beginning of the path
Output
- Admissible (vmin,vmax) at the end of the path
Pham, Caron, Nakamura RSS 2013
Admissible Velocity Propagation (AVP)
SLIDE 20
Based on Bobrow algorithm Implemented in TOPP
Admissible Velocity Propagation (AVP)
SLIDE 21
Planning using AVP
SLIDE 22
Planning using AVP
SLIDE 23
Planning using AVP
SLIDE 24
Planning using AVP
SLIDE 25
Planning using AVP
SLIDE 26
Example : Non-prehensile transportation
No need to design specific grippers Save time on grasp/release Use friction
SLIDE 27
Approach to integrate dynamics into motion planning Can be built upon existing sampling-based planners Negligible overhead over quasi-static planning Source code available https://github.com/quangounet/TOPP Current work
–
Liquid transportation
–
Humanoid robot
–