CPS Graduate Seminar Student Presentations
Mobile Robot Navigation with Neural Maps Michail G. Lagoudakis Department of Computer Science Duke University (Center for Advanced Computer Studies) (University of Southwestern Louisiana)
Motivation 4 Autonomous Robots –Planetary Exploration –Service Robots –Inspection Robots 4 Biologically-Inspired Robots –Animal Learning and Behavior –Neural Network Models
Talk Outline 4 Background –Mobile Robot Navigation –Neural Maps 4 Navigation with Neural Maps 4 The Polar Neural Map 4 Implementation on a Nomad Results 4 Conclusions
Mobile Robots RWI B14RWI B21 Nomad 200Nomad 150Nomad XR4000
Mobile Robot Navigation 4 Global –Map-Based –Deliberative –Slow 4 Local –Sensory-Based –Reactive –Fast
Navigation Subproblems 4 What should I remember? Cognitive Mapping 4 Where am I? Localization 4 Where should I go? Path Planning 4 How can I go? Motion Control
Animal Navigation
Robot Navigation
Navigation Landscape
Neural Maps 4 “A localized neural representation of signals in the outer world” [Amari] 4 Hopfield-type Neural Networks 4 Topologically Ordered Units
Neural Map Property 4 Amplification through Self-Organization
Neurons 4 Non-linear Processing Units 4 Non-Linear Dynamics
Path Planning with Neural Maps
Network Topology
Path Planning Example 1
Path Planning Example 2
Problem? 4 Global Information?
A “Bad” Idea
A “Good” Idea
The Polar Neural Map 4 Represents the local space. 4 Resembles the distribution of sensory data. 4 Provides higher resolution closer to the robot. 4 Conventions: –Inner Ring: Robot Center –Outer Ring: Target Direction
Example
Example (...continued)
Boudreaux (Nomad 200) 4 Nonholonomic Mobile Base 4 Zero Gyro-Radius 4 Max Speeds: 24 in/sec, 60 deg/sec 4 Diameter: 21 in, Height: 31 in 4 Pentium-Based Master PC 4 Linux Operating System 4 Full Wireless 1.6 Mbps Ethernet 4 16 Sonar Ring (6 in in) 4 20 Bump Sensors
System Architecture
Results
Results (…continued)
Robot Movie Enjoy some robotic video footage!
Contributions 4 Neural Maps for Fast Path Planning 4 The Polar Neural Map 4 Implementation on a Real Robot 4 A Complete Local Navigation Scheme
Future Work 4 On-board Code Execution 4 Polar and Logarithmic Map 4 Global Navigation 4 Neural Map Self-Organization
More Information 4 M.Sc. Thesis, Poster – 4 IEEE ICRA ’99 Paper –
Thank You TheEnd The End! © CPS 300 Inc. Special thanks to The Robotics and Automation Lab at USL Prof. Anthony S. Maida Prof. Kimon P. Valavanis Prof. Bill Z. Manaris