|
NASA
Ames Phase II SBIR for “A Hybrid End-to-End Architecture that uses
Model- and Rule-Based Inferencing” Columbia, Md., May 10, 2002 -- Interface & Control Systems, Inc. (ICS) has developed and prototyped, as part of the Phase-I SBIR, a software architecture that demonstrates the feasibility of integrating sequence, rule, and model-based autonomous reasoning components into a cohesive command and control, and fault monitoring and recovery system. During Phase II, ICS will formalize a UML modeling approach for modeling autonomous systems that is independent of any particular autonomy component, but which provides the foundation to auto-generate the run-time artifacts required by the autonomy components. Translators to generate Livingston state models and SCL scripts and rules from a UML model are being developed as a specific implementation of this approach. ICS has also secured a flight demonstration of this integrated SCL/L2 architecture as part of the AFRL’s TechSat-21 Cluster Manager software experiment. The development and integration of the flight components will be completed for this demonstration. The Cluster Manager experiment will perform autonomous operations and simulated fault monitoring and recovery for a small cluster of spacecrafts flying in close formation. ICS, which provides SCL to the industry as an open-source software product, also plans to continue to work closely with Ames to make L2 a commercially available, open-source, software product. ICS believes that these two technologies are highly complementary and provide significant benefits towards the goal of achieving completely autonomous systems. The key to success is in the integration of these technologies into a flight proven architecture, supported by appropriate modeling and ground system tools. Several commercial and government customers have expressed a strong interest in the aforementioned technologies to increase autonomy and automation of their missions. ICS is working with Lockheed Martin and Grumman on the SBIRS-High Ground System. The ground system, which uses SCL for fault detection, must support the payload flight software in providing system level anomaly detection and resolution. System-level state identification/estimation is essential and Livingstone’s capabilities in this area would greatly augment SCL’s sequence-based executive and rule-based inference engine. ICS is also currently under contract with Honeywell to support the Second Generation Reusable Launch Vehicle Integrated Vehicle Health Management (RLV2 IVHM) system. ICS is also seeing an increased demand for these technologies in non-aerospace areas such as transportation and robotics. The integrated, event driven, model based system provides a proven, cost effective solution to a wide range of control system problems and market segments. - end - Contact information:
 |
