(30.8MB) Some experimental results from a test of an algorithm to find the safest path through a field where chlorine gas is slowly drifting across the field (simulated by using a flashlight to activate the sensors). The skulls are the sensors, the gas appears as red areas of interpolated sensor data, and then a safest path is computed by the sensors for a person to exit the field. This was run on a real testbed of sensors.
(13.9MB) A panning shot of the pile of rubble at the Lakehurst Air Force base. This rubble is used for training search and rescue teams. I went there to deploy a sensor network on the rubble as a preliminary investigation to see what kind of data we could get.
(10.7MB) A crane robot, suspended by cables from the ceiling attached to winches scanning back and forth across a sensor network to calibrate it.
(8.4MB) Video, simulation output, and animations of data showing some of our work with electronically herding cows using sounds.
(16.2MB) Video of an autonomous robot helicopter flying over a sensor field in Australia. The sensors are next to the widely dispersed red flags on the ground.
(10.7MB) Close up video of the robot helicopter.
(9.3MB) Panning shot of the sensor field in Australia. The sensors are placed on top of plastic flower pots!
(1.9MB) The sky in Australia. Can't do science all the time!
(31.7MB) Video showing an autonomous robot helicopter auto deploying sensors in a pattern at USC in California.
(31.6MB) Video showing an autonomous robot helicopter auto localizing sensors so the sensors know where they are.
(17.6MB) Video showing an autonomous robot helicopter auto repairing the sensor field. The sensors have computed where the areas are that they are unable to talk to each other via radio and have reported that to the robot which then deploys more sensors in those areas so that coverage is complete.
(29.9MB) Video taken from a video camera attached to the robot helicopter.