Software Agent-based Framework Supporting Autonomous and Collaborative Sensor Utilization (AAMSRT)

about the world and its local environments prevails overwhelmingly due to the rapidly growing availability and heterogeneity of sensors and sensor platforms (e.g., various advanced imagers installed on different Low Earth Orbit (LEO) satellites providing observations of an event from different views). The next logical step is to synergize these assets in order to provide a more complete and coordinated view of user interested events. Currently, users that require sensor measurements have to be able to determine which sensor platforms and sensor capabilities should be used to fulfill their observation requests. This condition incurs the dependency of an expert to intermediate and plan the observation, significantly reducing the utility of an already large and growing sensor capability. Our approach was to develop an intermediary step that translates user's domain specific requirements into domain independent observation requests. The resultant domain-independent formulation may be viewed as coordinated service planning and execution employing heterogeneous sensor assets. In a sense, we have proposed a commoditization of sensor capabilities as well as a translation layer that determines the commodity services required to satisfy a domain specific request. In addition, we have implemented an agent-based autonomous multiple sensor re-targeting framework (AAMSRT) that supports open, heterogeneous and dynamic environments to negotiate and track the execution of these services. This innovative approach involves the abstraction of sensor assets as software agents that encapsulate sensor services allowing for easy coordination between multiple providers, contractual support and business rules using a market approach. The tool provides a web-based interface to allow a user to formulate a service request, as well as monitoring execution tracking and replanning capabilities. In summary, our AAMSRT framework enables coordinated employment of heterogeneous sensor assets on a service architecture for building business applications, e.g., earth modeling and observation. This paper addresses the multi-sensor retarget problem at a high level and provides a complete solution.