Airport Ground Operations Meets Computer Vision

Computer Vision Will Revolutionize Gate and Asset Management While Improving Safety

Several academic and industry journal articles have discussed the application of computer vision using artificial intelligence for airport operations. They usually center around improved security screening of passengers and luggage. Of the dozen or so articles I found, none discussed computer vision to track and monitor ground operations to increase the velocity and throughput of gate turnaround times and identify potential problems in order to prevent delays and increase safety.

The diagram below from the Boeing Corporation illustrates just how many vehicles and pieces of equipment are involved in servicing an aircraft.

  1. Tow Truck/Tractor
  2. Electrical Power
  3. Galley Service Vehicle 1
  4. Galley Service Vehicle 2
  5. Baggage Tags
  6. Baggage Carts
  7. Bulk Cargo Loaders
  8. Jet Bridge
  9. Lavatory Service Vehicle
  10. Cabin Cleaning Vehicle
  11. Potable Water Vehicle
  12. Conditional Air Vehicle

There is typically a mix of fixed systems and mobile equipment that are used to service aircraft. Mobile equipment offers greater flexibility than fixed equipment but comes with higher rates of accidents and other incidents. Accidents can include equipment impacting with the aircraft, which takes the aircraft out of service.

Studies have shown that ground servicing of a commercial aircraft can run from 20 to 60 minutes with upwards of 20 distinct activities, many of which occur simultaneously. This table is an example of the time to service a wide body jet.

With the level of commercial traffic having doubled in the last 12 years, and the inability to easily increase terminal capacities, the demands to improve gate turn rates can be expected to increase.

Gate management can be complicated by the use of third-party contractors. This is especially the case in the EU. In other words, there may be several different contractors involved in managing the gate service. Even if one organization is managing all activities, there is still the issue of so many activities in close proximity, occurring over a short period of time and often times simultaneously.This is where computer vision can help in a significant way. Cameras and machine learning can monitor and analyze all the vehicles and activity around the aircraft in real time. Alerts can be created based on user-defined rules such as identifying when the catering service has not arrived after a prescribed amount of time, or a service vehicle that has exceeded a defined timeline to service the aircraft, or equipment has been left in non-compliant locations around the aircraft creating a safety hazard.

With new insights that come with computer vision, gate management is more efficient resulting in faster gate turns, fewer delays and greater safety. In addition, the ability to archive months’ worth of computer vision data provides both quantifiable and qualitative datasets that can in turn be used to facilitate continuous improvement efforts.

Anthony Tarantino, PhD
Adjunct Professor, Santa Clara University – Operations and Finance
Six Sigma Master Black Belt, Certified Scrum Master, CPIM (APICS), CPM (ISM)
Senior Advisor to Atollogy