Research
Path Planning for Redundant Manipulators in Highly Cluttered Environments
Key Challenges
High-dimensionality of the search space
Novel Approach
A general bi-directional tree-search framework
The framework allows balancing reliance on workspace cues and configuration space sampling
Key Results
Very low path planning failure rate
Improved path quality
User-Guided Path Planning for Redundant Manipulators in Highly Cluttered Environments
Key Challenges
High-dimensionality of the search space
Novel Approach
A general bi-directional tree-search framework
The framework allows robot operators to provide high-level guidance during the search process
Key Results
Very low path planning failure rate
Improved path quality
Trajectory Planning for Robotic Boats in Complex Environments
Key Challenges
Navigation through wave-fields of traffic vessels
Robot motion uncertainty
Perception uncertainty
Novel Approach
Incorporate wave-fields, motion uncertainty, perception uncertainty in the trajectory planning problems
Key Results
Low collision rate
Improved trajectory execution time
Trajectory Planning for Unmanned Ground Vehicles in Complex Environments
Key Challenges
Navigation over uneven terrain (slopes)
Presence of dynamic obstacles
Perception uncertainty
Novel Approach
Use approximate acceleration/deceleration model of the UGV
Regulate velocity along curve to avoid collisions
Key Results
Low collision rate
Improved trajectory execution time