Mapping System of Water Pollution by Autonomous Fish Robots

We propose a water pollution source mapping system based on ubiquitous sensor networks that fish robots search the source of water pollution autonomously. To verify the effectiveness of a water pollution source mapping system, we made a model water pool in which a fish robot pursues infrared measurements of reflection to various colors at the bottom of the pool. The model of water pollution consists of various color maps, an LED emitting infrared, and a photo-transistor sensing the same wave length light. Different color has different reflectivity for the same light source; therefore this simulation model imitates the diffusion of water pollution sources. For real tests, an LED and the photo-transistor can be replaced with particular water pollution sensors. Water pollution maps can be composed based on the known mote location information by identifying motes. To get improved position data between the motes, directional sensors such as magneto-resistive sensors and accelerometers are applied to compensate noise components due to waves and the swing actions of the tail fin. Fish robots obtain simulated pollution data in terms of reflectivity autonomously. Key-Words: Fish Robot, Pollution Map, Ubiquitous Sensor Networks, Noise Reduction