Continuous visibility into aircraft turnaround operations.
Founding Minds developed AutoATOps, a computer vision system that monitors aircraft turnaround activities using an airport’s existing camera feeds. It detects and timestamps ground-handling events automatically, reducing dependence on manual activity logging and giving operations teams a more timely picture of turnaround progress.
The situation
Aircraft turnaround is the sequence of tasks between an aircraft landing and being ready to depart again. The tasks run in a specific order and to precision time scheduling.
In most airports, the record of what happened is kept by hand. The people performing ground operations log each event into the system on handheld devices, while doing the job the event describes. When the ramp is busy, logging competes with the work, so the record can lag or be inconsistent, which limits visibility into turnaround progress.
What made it hard
Working with existing infrastructure.
The solution had to use the surveillance cameras already in place, with no purpose-built cameras, ultrasonic sensors, or special-purpose devices on the stand, and without a significant overhaul of the software that the ground crew and the airline already run.
Integrating with different airport operations.
Operational processes and tools vary from airport to airport, and so does the way data is shared between airline and airport. When ground operations are outsourced, the handling agent can change over time. There is no single integration to build, so event logging had to accommodate those variations.
Understanding activity in difficult visual conditions.
An aircraft stand is an uncooperative place for a machine to read. Vehicles occlude each other, ground crew move in and out of frame, lighting runs from daylight to floodlit night, and weather varies.
Identifying operational events accurately.
The timing of each event matters. The system has to recognize specific activities and timestamp them accurately enough for operational monitoring, avoiding both false detections and missed activities.
What was built
AutoATOps generates software triggers for turnaround events, in three connected parts.
Visual monitoring
Takes live feeds from surveillance cameras covering the gate, the boarding area, and the apron.
Event detection
Applies computer vision to track aerobridges, fuel trucks, baggage trucks, and belt loaders, keeping a continuous picture of what is where in relation to the aircraft, and flags each turnaround event as it occurs.
Timestamped event records
Generates timestamped software events that support operational monitoring and the workflows the airport and airline already use.
- Aircraft parking
- Aerobridge connection and disconnection
- Cargo door open and close
- Cargo onloading start and stop
- Cargo offloading start and stop
- Catering loading start and stop
AutoATOps automates observation and event recording while leaving operational decisions with the ground crew.
For operators with stricter infrastructure requirements, the system can run on a dedicated on-premises server.
What changed
A more timely record of turnaround activity
AutoATOps automatically detects and records supported ground-handling events. This reduces dependence on manual event entry and gives operations teams a more consistent view of turnaround progress.
The engineering decision that mattered
The system was designed to observe existing operations rather than change how the ramp works. By automating event recording instead of operational decisions, AutoATOps addressed a difficult visibility problem without adding a new layer of control.