Make robot learning practical by integration with humansupplied knowledge

1
Intelligent Adaptive Mobile Robots

• Make robot learning practical by integration with
  human-supplied knowledge
• Solve huge problems through continuous reframing
• Achieve robustness by explicit uncertainty
  modeling and information gathering

1. Develop learning architecture 2. Implement
and test optical-flow-based obstacle avoidance 3.
Demonstrate learning amplification of human
expertise in simple domain 4. Develop and
implement learning algorithm for huge uncertain
domains 5. Integrate and demonstrate learning,
planning, navigation in complex domain

• Fast development of robust, agile mobile robot
  software
• Robots capable of extended operations in complex,
  uncertain multi-objective domains
• Adaptive navigation and mapping based on passive
  visual sensing

1
3
2
4
5
7/99
7/01
7/00