Creating a plane tracking camera

One of the major projects I'm working on this year is creating an autonomous, AI powered, plane-tracking camera that live-streams landing aircraft at a major airport. This project is ongoing and wildly complicated in execution.

The Idea

I had the idea about 10 years ago, when I was deep in flight training and wanting to review my own landings. It crossed my mind, but I quickly dismissed it as a "dumb idea" which "nobody would ever want". Fast forward to 2025, after discovering that other things like this actually do exist, I thought it would be a fun technical challenge.

Progress

Progress so far has been good, but it is not without its challenges. I started by creating a software solution that is able to pull ADSB data directly from a receiver, display it on a map, and allow for selection of particular aircraft to track.

I then created a module that calculates the azimuth and elevation needed for a camera to point in the direction of the aircraft, given that aircraft's current position, a time horizon (since the camera will have a delay for moving), and the latitude, longitude, bearing, and altitude of the camera itself.

Testing At A Small Airport

I called a bunch of small airports, and was eventually able to negotiate a time to come test out my device on airport grounds, at a busy uncontrolled airport. The device at the time relied heavily on ADS-B transmissions, but unfortunately, private aircraft in Canada aren't always outfitted with these devices, and most aircraft in the pattern were not transmitting.

Back To The Drawing Board

I then went back to the drawing board, and decided to implement a half-ADSB half-PID tracking system, using YOLO. The idea was to use ADSB (or a fixed point on the camera) to locate it initially, and then use an object recognition algorithm like YOLO, and PID control to keep that aircraft in focus as it conducts its landing.

Testing At A Large Airport

I then asked pretty much everyone I knew who owned office space close to Billy Bishop Aiport, if I could set up there for a week and test out this camera on large landing aircraft. The tests were mixed - I was able to keep the aircraft in frame somewhat, but the speed at which they crossed the frame made it difficult to keep them in frame.

Negotiating Lease Space

Putting the technical work on pause for a moment, I started turning my efforts towards trying to secure a lease for roof space on a private business close by a few target airports. This has been the largest challenge of the project so far, and has been my entire focus since then. I know I can technically complete this project, but the challenge of trying to find a partner who wants to be a part of a novel project like this is overwhelming - but I'm keeping my hopes high!