Daher, the French maker of the TBM and Kodiak series of single-engine turboprops, recently held a unique contest in a search for new products users would like to see on the TBM series of very fast turboprops. Daher called the contest general aviation’s first hackathon or “General Aviathon.” The company chose three winning concepts that proposed highly imaginative solutions to further increase flight safety for the company’s TBM turboprops while also enriching the in-flight experience for passengers. The hackathon focused on two TBM operational themes: advanced features and capabilities to make the aircraft more connected, intelligent and autonomous; and innovative solutions to further refine the experience of piloting and using the TBM.
One of the winners was the Aviation Voice Recognition System (AVRS), proposed by Adam Kaplan and Ben Krosner, two engineers at the Massachusetts-based Scientific Systems Company, Inc. Their AVRS software package is capable of parsing and understanding instructions from air traffic controllers to make changes in the aircraft’s flight profile (such as climbing, descending and turning), which would be implemented subject to the pilot’s approval.
Leveraging the Microsoft Azure public cloud computing platform for acoustic modeling, the AVRS’ development would focus primarily on an aviation-centric language model created with real-world audio examples from air traffic control. By processing ground-truth flight data from the aircraft’s flight deck avionics (including location, altitude, and flight plan), the AVRS’ computations could be handled onboard, removing the need for cloud computing.
Also selected as a safety-related winner of the General Aviathon was an idea for virtual pilot monitoring on the TBM, providing verbal guidance based on images processed from cockpit cameras to assist the pilot in flying the aircraft. The “TBM Monitoring Pilot” proposal was submitted by Vijay and Meera Singh–husband and wife entrepreneurs who own and fly a TBM 900 aircraft based at Somerset Airport in New Jersey.
As TBM aircraft are typically operated by a single pilot, the “TBM Monitoring Pilot” would offer assistance by using a processor to perform object recognition, and then provide audio guidance to the pilot’s headset based on a deep-learning training set. Advantages of this proposal include the use of off-the-shelf processors developed for autonomous cars, as well as working with a finite set of objects to be recognized for rapid training of the system using artificial intelligence and machine learning.
The hackathon’s third winning entry, “Avia-X,” proposed a real-time passenger experience system conceived by four students at the University of California-Berkeley: Alan He, Frank Ge, Gary Yang and Ken Guan. Avia-X would enhance the on-board environment with pre-set themes and emotion recognition for light and music, potentially expanding to include an olfactive experience. “Avia X” could be managed via the “Me & My TBM” cloud-based application for Android and iOS devices, which has been developed by Daher to support TBM owners and operators to track aircraft operations and maintenance.
The “General Aviathon” challenge was originally organized by Daher’s Silicon Valley-based innovation program, Armstrong by Daher. Winning ideas were selected this month following a competition that attracted 25 proposals from the United States, China, India and Mexico.The winning concepts were chosen by a jury of representatives from Daher and its hackathon technology partner, Microsoft, as well as from San Francisco-area startup companies, accelerators and the academic community. The pitches from 11 finalists were made during an event at the Hiller Aviation Museum in San Carlos, California, with more than 100 people in attendance. Each winner received a cash prize of $5,000. Each also has the potential of partnering with Daher in further development/deployment of their solutions.
