While IBM is applying its Watson system to a variety of airline problems, former IBM aerospace executive Anant Sahay started his own firm, 1Ansah, to apply natural language processing and machine learning to help mechanics do faster, smarter trouble-shooting. The 1Ansah tool, AMRIT, is already working for Airbus helicopter techs in Australia, and Sahay is now setting up pilot test of AMRIT with SIA Engineering Company.
AMRIT enables engineers and techs to analyze technical publications, maintenance logs, MRO systems and sensor data in a single source of easily searched content that presents all the information in indexed, structured and ordered categories. Then AMRIT’s advanced intelligent assistant advises on relevant trouble-shooting and repair options and ranks them according to past success rates.
AMRIT won an innovation award from Frost & Sullivan and has received new investment from the Datt Group of private investors. At Airbus Helicopter, two new machine learning products have been added, one more is in development and three more are in planning. One new product automatically classifies technical documents based on content, reducing the time and effort to update the security system. Another new function revamps document management to make it intuitive and intelligent. “In the future we are looking at working with data collected from sensors fitted in helicopters and intelligent reporting,” Sahay says. The Datt investment will enable 1Ansah expand AMRIT capabilities by adding deep learning and pattern recognition, that is, item recognition using augmented reality.