Lichten Award Paper: Automated Optical Rotor Blade Tip Clearance Tracking Using Artificial Intelligence Algorithms

Analytical methods used to determine weapon projectile trajectory relative to helicopter rotor blade tips during most conceivable maneuvers tend to provide insufficient safety margins to accurately define a safe firing envelope. The flight test team at Sikorsky, a Lockheed Martin Company, was given the challenging task of measuring the weapon-to-rotor-blade-tip clearance during these maneuvers. The first-generation methodology was entirely manual using a flat vertical projection screen tangentially aligned with the rotor tip path plane to visualize the intersection between the weapon's projectile and blade tip path. While the first flight test program utilizing this technique was successful in its own right, by the time a second opportune flight test program was initiated, software tools had been developed to help automate the process, drastically increasing the amount of data that could be used to correlate with other aircraft state parameters. During the data processing, a training set was created that was used to build an Artificial Intelligence (AI) capable of performing the same task while bringing the processing time down from 4-5 frames per second (fps) to 130 fps. This enables real-time AI inference to be taken from digital video cameras on the helicopter and clearance measurements real-time processed and recorded to the instrumentation system or displayed for the pilot's awareness.