Dr. Kanade’s research in vision has led to several advances in sensory equipment for autonomous mobile robots, that sense and navigate their environment free of human intervention. One project was the development of the Autonomous Land Vehicle in a Neural Network, or ALVINN [46pp]. This system provides a robotic learning mechanism that allows a vehicle to be trained to drive independently of human control. The successful creation of a chip for high speed processing of images and computation required for successful navigation, known as ALVINN On a Chip [7pp], further enhanced the project.
Dr. Kanade has also worked extensively on navigation and interpretation [34pp], closely related to the work for the ALVINN Project, which focuses on methods to combine computational data and image processing, creating better visual sensing methods. Kanade’s research in autonomous robotics also led to the development of sensory equipment for identification of rock samples [1p] for planetary explorers on Earth and in space.
Dr. Kanade is currently collaborating on the development of an autonomous helicopter [1p] at Carnegie Mellon, which would allow for helicopters to undertake tasks that are too dangerous for human pilots to execute. Search and rescue, fire prevention, and intelligence gathering could now be preformed independent of pilots, and greatly enhance the versatility of these aircraft. For further information visit the helicopter lab’s homepage at http://www.cs.cmu.edu/afs/cs/project/chopper/www/.
J. Dustin Williams, University/Heinz Archivist, Carnegie Mellon University