WTC 2005 Year End Update

Maritime search and surveillance-type missions typically require heavy human involvement. A set of unmanned aerial vehicles (UAVs) has the potential to provide a “sensor network” to greatly increase efficiency and effectiveness of these surveillance type missions. However, limited autonomy is still a bottleneck to networked UAV applications. High operator involvement is required in logistics and operation, for example, distributing assignments such as which regions to search and coordinating subsequent sensor measurements. In a noisy environment, it becomes difficult for a human operator to classify sensor readings and assign confidence in these readings. Determining regions of high target-location probability and coordinating nearby agents to converge on a particular spot while allowing other vehicles to continue searching is also difficult. Therefore, the primary limitation to concurrent operation of multiple vehicles remains lack of autonomy of these vehicles. The current work aims to develop target identification and searching algorithms for the Georanger UAV [1] for aero-magnetic surveying. A specific application is the detection of a submarine in littoral waters based on its magnetic signature. Eventually, these algorithms would be developed to operate with a team of heterogeneous vehicles. This team would be comprised of individual vehicles known as agents. Each agent could have different capabilities and sensors which dictates that algorithms be easily adaptable to accommodate these differences. A summary of the completed, current, and future research activities with this project include: