Increasing ATM Efficiency with Assistant Based Speech Recognition

Abstract

Abstract—Initiatives to integrate Automatic Speech Recognition into Air Traffic Management (ATM) exists at least since the late 90s. Some success to replace pseudo pilots have been reported, but its integration into controller assistant tools is missing. German Aerospace Center (DLR) and Saarland University developed Assistant Based Speech Recognition (ABSR) enabling command recognition rates better than 95%. However, good recognition rates are no convincing argument for decision makers. Therefore, we conducted an ABSR validation study with eight air traffic controllers to quantify the benefits with respect to workload and efficiency. The study validates that ABSR does not just reduce controllers’ workload, which would already be a lot, but this paper presents that ABSR significantly increases ATM efficiency. Fuel reductions of 60 liters (16 gallons) per flight and a throughput increase by two arrivals per hour are possible.