1. Introduction
There are two kinds of planning applications at NASA: (1) those for autonomous spacecraft and (2) those for missions. Each of these types of planning applications is subject several impediments including the following three major difficulties:
- Exceptions and unforeseen circumstances occur
- Finding a correct and high quality plan is a difficult search problem (complex problems and large planning spaces)
- Constraints appear, evolve, and disappear over time.
The planning spaces that NASA deals with for areas such as spacecraft operations and Mars rover navigation are quite large. An exhaustive search of planning spaces this large is clearly impractical for dynamic planning. Current approaches to such applications emphasize either the construction of large comprehensive plans that seek to encompass as many contingencies as possible [Washington et al., 1999], or the construction of reactive systems that use a greedy or serendipitous approach to planning. Constructing large plans is difficult, because exceptions cannot always be foreseen and opportunities are often missed. Moreover, once a large plan is enacted, it is difficult to monitor and modify.
We propose to develop and investigate a normative reasoning system for planning and plan execution. The core hypothesis of our proposed research is that highly distributed planning based on normative reasoning can be used to produce coherent behavior in very dynamic environments. We will investigate this premise in the context of planetary rover navigation and task planning.
The key technology areas covered by this proposal are:
- highly-distributed planning
- agent-encapsulated decision networks
In addition, since we are addressing rover navigation, we will develop agent-based subsystems for handling robotic vision and sensor data fusion.
The proposed research specifically targets priority research areas for the AI Group at the NASA’s Jet Propulsion Laboratory, including autonomy, automated planning and scheduling, control architecture and executive, and navigation and control. It also targets information technology which is a priority technology are for South Carolina. There are significant benefits to the state of South Carolina from the proposed research. First, it contributes to the priority research area of information technology by developing the infrastructure for normative reasoning and distributed agent technology. The development of efficient algorithms will make it possible to transition this technology from academia to industry. Second, the majority of the requested funding is for graduate students and a postdoctoral associate. The benefit to the state is the development of a trained workforce in cutting edge technology. Finally, this proposal research will foster a long-term collaboration between researchers at the University of South Carolina and NASA.