|

Program Summary
The objective of the RPA ATD was to apply artificial intelligence and state-of-the-art computing technologies to manage and integrate next-generation mission equipment and battlefield information to enhance lethality, survivability, and mission effectiveness of combat helicopters. The primary element of the RPA system is the Cognitive Decision Aiding Subsystem, which performs battlefield situation assessment, planning, and cockpit information management. As a proven application of associate system technology, RPA has become essential to forming the requirements for future warfighting commanders and the combined arms team.
The Rotorcraft Pilot's Associate (RPA) Program focused on creating tactical Cognitive Decision Aids for reconnaissance and attack helicopter cockpits, like that of the AH/64-D Apache shown above. These Decision Aids act as the Pilot's "associate" by computing and displaying tactical information, just-in-time, tailored to the Pilot's needs and preferences. The RPA system was successfully flight demonstrated onboard an AH/64-D in August 1999.
ATL Role and Enabling Technology
RPA's Battlefield Assessment functionality requires a real-time view of the locations and identities of all battlefield air and ground entities. ATL's multi-sensor, multi-target onboard/offboard Data Fusion Subsystem drives Battlefield Assessment with its once-a-second reporting of the fused friendly and hostile tracks correlated from the more than 14 onboard and offboard information sources shown above. Data Fusion was determined to be a significant technology enabler for reducing pilot-intensive activities in the cockpit and for improving situation awareness.
Key Program Challenges
Specific challenges that ATL encountered and resolved during the RPA ATD include:
Estimating the performance of RPA onboard and offboard information sources for the 2010 time frame.
Fusion of both kinematic and classification information in real time.
Applying a real-time algorithm to the typical fusion bottleneck of Assignment using the JVC algorithm.
Combining information with incompatible accuracies without incurring large error versus ground truth (a 3% assignment error was required, ATL achieved <1%).
|