Neural Network Limit Avoidance System for Rotorcraft

Abstract

Barron Associates is designing and developing systems that help reduce pilot workload and help pilots avoid exceedances of operating limits under adverse conditions. Barron Associates’ Neural Network Limit Avoidance System predicts impending exceedances and warns pilots before they occur, allowing them to take corrective action to avoid the exceedance.

Problem

Piloting a helicopter under normal circumstances is a highly demanding task, and under adverse conditions is highly demanding of the pilot to the degree that s/he cannot simultaneously achieve all of his goals. Increased exceedance of the aircraft’s operating limits may occur in such situations, leading to increased maintenance requirements and in extreme cases the possibility of system failures. There is a need for systems that help reduce pilot workload and help pilots avoid exceedances of operating limits.

Solution

Barron Associates, working with Matthew Whalley at NASA Ames Research Center, developed a system that predicts near-future exceedances of operating limits and alerts the pilot to these impending exceedances through force feedback cues on the control sticks and visual cues on the heads up display. The system was evaluated in the vertical motion simulator facility at NASA Ames with favorable results.