Intelligent Information Processing for Enhanced Safety in the NAS

Our Phase I work focused on how improved information flow between actors in a flight deck environment can improve safety performance. An operational prototype was developed demonstrating how the Intelligent Information Processing System (IIPS) will operate in actual accidents/incidents.

For Phase II, we propose the following operating environment extensions from the flight deck environment: NextGen scenarios emphasizing interactions with air traffic controllers operating in fast paced, increased volume of manned and autonomous traffic; UAV operations emphasizing introduction of UAVs into the NAS, transition to autonomy and fully autonomous operations; and IIPS in flight training environments both simulated and airborne.

We also propose an extension to the manner in which conditions were developed in Phase I. Conditions were developed using post analysis of accidents and incidents. The error chain of events was identified, information necessary to prevent the event was identified, and finally, a condition developed that detected the circumstances for a possible safety failure so that a notification could be transmitted to the actor who would then take the appropriate action to break the error chain. This paradigm of condition development can be characterized as reactive. With the NAS moving into a state of flux with the integration of UAVs and general increased traffic volume, reactive safety may not be acceptable. In order to continue the steadily improving safety record of aviation, a more proactive approach must be considered. We propose the use of a classical rule-based expert system and other artificial intelligence approaches that can make inferences of possible unsafe conditions using a temporal knowledge base populated by propositional statements generated by IIPS information sources.