Christian Plaunt

A hybrid procedural/deductive executive for autonomous spacecraft

The New Millennium Remote Agent (NMRA) will be the first AI system to control an actual spacecraft. The spacecraft domain places a strong premium on autonomy and requires dynamic recoveries and robust concurrent execution, all in the presence of tight real-time deadlines, changing goals, scarce resource constraints, and a wide variety of possible failures. To achieve this level of execution robustness, we have integrated a procedural executive based on generic procedures with a deductive model-based executive. A procedural executive provides sophisticated control constructs such as loops, parallel activity, locks, and synchronization which are used for robust schedule execution, hierarchical task decomposition, and routine configuration management. A deductive executive provides algorithms for sophisticated state inference and optimal failure recovery planning. The integrated executive enables designers to code knowledge via a combination of procedures and declarative models, yielding a rich modeling capability suitable to the challenges of real spacecraft control. The interface between the two executives ensures both that recovery sequences are smoothly merged into high-level schedule execution and that a high degree of reactivity is retained to effectively handle additional failures during recovery.

Run-time satellite tele-communications call handling as dynamic constraint satisfaction

The next generation of communications satellites may be designed as a fast packet-switched constellation of spacecraft able to withstand substantial bandwidth capacity fluctuations due to causes ranging from unstable weather phenomena to intentional jamming. Scheduling and servicing call requests in such a dynamic environment requires real-time decisions with regard to allocation of resources including bandwidth, call routing and load balancing. In this paper, we present a general satellite communication scheduling domain, and describe a working implementation of an autonomous system for handling such dynamic scheduling problems. The solution approach is drawn from the area of dynamic constraint satisfaction problems (DCSP), which generalizes these and many other dynamic scheduling problems. We adapt DCSP techniques to the satellite communications domain, in particular solution repair and optimization by gradient-climbing. These reasoning methods respond to changes in the problem specification by repairing the current solution. As a result, they are anytime algorithms which can trade run-time efficiency for solution quality. This approach supports dynamic call requests; negotiation and fulfillment of prioritized Quality of Service (QoS) contracts; graceful degradation in the presence of dynamic call traffic, changes of priority schemes, and environmental conditions; and optimization of geometrically constrained resources.