Sensor-based Robot Motion Planning
- An implementation of ER1 robot
The design of a navigation system for mobile robots. To safely move a robot in a given environment there are at least three issues to concentrate on: (1) the technique to generate the collision-free motion commands - usually is a collision avoidance technique. (2) the specific type of robot. The robot's characteristics (shape, kinematics, and dynamics) establish requirements and constraints over the collision avoidance methods. (3) the fact the robot is required to move in a given environment. This imposes into the navigation system other requirements and constraints that relate with the environment nature: the environment can be highly dynamic, non-predictable, unstructured, complex, dense, and cluttered...
This final project analyzes the requirements, influences, and relationships among these issues, and try to presents some innovative solutions for all these issues in a robust navigation system based on ER1 robot platform.
1. Goal (possibilities)
(1) Sensor coverage path planning;
(2) Depth from single camera;
(3) Dynamic obstacle avoidance.
2. ER1 Robot
"The ER1* is the first robot with professional-level robotics software technologies and industrial grade hardware designed for enthusiasts like you who are interested in state of the art technology that takes advantage of your technical skills and imagination.
The ER1 has computer vision, hearing, speech, networking, remote control, email, autonomous mobility, gripping, and IR sensing capabilities - all brought together with an open software system and a reconfigurable chassis.
Accessories are also available for purchase, helping you to customize and expand your robot in hundreds of ways."
---- Quoted from http://www.evolution.com
*ER stands for Evolution Robotics, for more details please refer to the commercial website.
ER1 video demos: (Requires Windows Media Player 9.0)
(1) The ER1 In Action
(2) The Robot Control Center Software
(3) Assembling the ER1
3. Technical Papers
 Hector H. Gonzalez-Banos, David Hsu, Jean-Claude Latombe, Motion Planning: Recent Developments
 Je Michels, Ashutosh Saxena, Andrew Y. Ng, High Speed Obstacle Avoidance using Monocular Vision and
Reinforcement LearningComputer Graphics, Computer Science Department, Stanford University, Stanford, CA 94305 USA.
 Ashutosh Saxena, Sung H. Chung, and Andrew Y. Ng, Learning Depth from Single Monocular Images,Computer Science Department, Stanford University, Stanford, CA 94305 USA.
4. Current Status