Collect and process ADSB observations

I want to know more about the aircraft around me such as what a "normal" level of activity might be or the types of aircraft. Since I am only interested in aircraft which operate locally, and I don't need geographic displays etc. (already well supported by websites such as https://https://adsbexchange.com/).

In addition, I want to know more about the aircraft than is reported via ADSB (i.e. aircraft model and registration).

ADSBexchange offers an inexpensive REST API, but it only reports when an aircraft is in flight.

ADSB Collection runs on a standard Raspberry Pi 4 using a rtl-sdr running dump1090.

dump1090 exposes a small HTTP server and will return a json formatted report of all ADSB broadcasts.

Twice a minute, I ask dump1090 to report on current ADSB broadcasts. For each aircraft, I then request amplifying information from the ADSB exchange REST API. The collected output is written to json formatted file and uploaded to AWS S3 for later processing.

There can be multiple collection stations writing to AWS S3.

To prepare a rPi for mellow-hyena, you will need a rtl-sdr dongle (as above) and dump1090. AWS S3 is optional, but nice if you have multiple collection stations. Install the mellow-hyena source via git and tweak confifigure.yaml and crontab as needed. Collection is working correctly when you see json formatted files in the aws_export directory.

Processing consists of moving collected ADSB observations from S3 to a machine for parsing and loading into postgres.

For processing, create a postgresql database and populate with the script "add_schema.sh".

Then run the script "parser.sh" to parse collected files and load into postgresql.

mellow-hyena produces simple reports about upload and observations.