A Distributed Environment for Onboard Planning and Scheduling

Charles River Analytics

Cambridge, MA

Allows a constellation of spacecraft to function as a hierarchy of autonomous agents handling tasks from multiple sources.
Reduces a single complex global scheduling problem to parallel scheduling problems that are easily tractable
Minimizes inter-agent negotiation by executing a generous domain distribution algorithm during constraint management.

• Developed a vertically layered agent architecture incorporating a domain-independent constraint database, an inference layer incorporating domain-specific knowledge, and an interface for interacting with the outside world.
• Developed a 3-D visualization interface for the simulation.
• Demonstrated by analysis the benefits of distributed vs. central scheduling and of inter-agent negotiation.
• Illustrated ability to resolve difficult scheduling problems with formation based on Three Corner Satellite Constellation.

• Multi-agent architecture incorporated into Charles River Analytics’ Intelligent Agent Toolkit. Applicable domains include commercial satellite systems, supply chain management, and agent-based modeling and simulation.
• Planning and Scheduling incorporated into Charles River Analytics’ Intelligent Agent Toolkit. Applicable domains include transportation and inventory control.

• Multi-agent architecture employed in Air Force contract for a Multi Agent-based Satellite System for Information Fusion (MASSIF).
• Planning and scheduling engine used in Air Force contract for a Generic User-friendly Integrated Development Environment (GUIDE).

Return to NASA SBIR Success Listings