Steve LaValle is one of the pioneers of robot motion planning. I got the rare opportunity to talk with him in my podcast series on robotics.
Topics included his work on the Planning algorithms book, development of the rapidly-exploring random tree algorithm, minimal sufficient robot brains, doing motion planning research with his son, and his new life in Oulu, Finland. Please enjoy!
Content
00:40 Early Childhood
04:59 Getting into University
07:27 Life at University
10:13 Symbolic vs Visual Mathematics
11:11 Getting into Phd Program
12:12 How to get into MIT for students
13:09 Advice from Phd Advisor
14:17 First Paper
16:30 Options after Phd
17:47 Advice to Phd Students
19:19 Postdoc in California
21:46 Demo for Bill Gates
22:52 Becoming Professor at Iowa State
23:34 Developing RRT
26:21 Reason of longterm success of RRT
28:34 Sampling-based vs. Randomized
30:30 Could RoboticsGPT deprecate motion planning?
31:53 Value of motion planning research
33:07 Information spaces
34:10 Minimal sufficient robot brains
35:37 Writing "Planning algorithms"
40:33 Managing and organizing references
42:08 Deprecated parts of the book
43:16 Challenges in writing more books
44:19 Unpublished RRT book
44:28 Reading and favorite books
45:50 Reading Ayn Rand
46:53 On joining Oculus VR
50:51 Virtual reality research
52:24 Killer application for VR
52:54 Facebook buys Oculus
53:31 Control over time is most important
54:20 Meeting Brett Leonard
55:30 Huawei
57:16 On being successful
59:30 H-Index as performance indicator
01:00:16 Metrics for academic career
01:01:36 Time allocation
01:02:28 Tired phase of career
01:03:26 Advice to younger researchers
01:04:40 Fascination with Finland
01:07:46 Learning Finnish
01:08:12 What Steve misses about the US
01:08:55 Benefits from Saunas
01:10:18 Future aspirations of lab
01:11:21 Research with Alexander LaValle
01:13:15 Robotics applications for daily life
01:14:04 Gardner hype cycle of deep learning
01:15:18 Value of material science for robotics
01:16:12 Closing comments