BLE Beacons in Agrotech: A Hands-On Experience

As CTO Jan Palas explained, the company’s mission has long been to let farmers “spend more time in the fields and less back in the office.” But in 2018, AG Info made a pivotal expansion into telematics, driven by a simple question: can we automatically understand what work is being performed, on which field, by which machine?

Tracking machinery movement was only the first step. The real breakthrough came from identifying implements, because in agriculture, the implement defines the job. Whether a tractor is plowing, seeding, spraying, or harvesting depends entirely on which trailer is attached. Manual input wasn’t an option: “Drivers are humans, and humans make errors,” Palas noted. Instead, AG Info mounted Bluetooth Low Energy (BLE) beacons on trailers and used in-cab GPS devices to detect them. Combined with flespi's unified device infrastructure, this gave the company the foundation to build automated job-type intelligence across fleets.

However, six years of real-world use revealed that the simplicity of the idea hides enormous technical and behavioral challenges. Czech winters cripple batteries. High-pressure washing and constant vibration push device housings to their limits. Mud, sometimes half a meter deep, can fully absorb beacon signals. With conservative settings (max 10-second advertising intervals and 0 dBm power), AG Info could manage roughly two years of battery life, but customers resisted sealed batteries that required frequent replacement. “There is no perfect module,” Palas admitted. “Only one that is slightly less imperfect.”

Even more difficult were the messy realities of farm operations. Farmyards are dense clusters of beacons, overwhelming detection algorithms until AG Info introduced automatic yard-level exclusion zones. Trailer swapping without shutting off engines forced continuous signal monitoring and dynamic trip splitting. During harvesting, multiple machines in the same field all hear each other’s beacons, requiring strict rules prohibiting trailer switching inside field boundaries. And finally, the “tractor coffee break” - engines left running for long stretches at the edge of fields - necessitated discarding all beacon detections when a machine is stationary.

What seemed like a straightforward task: identifying which trailer is connected to which tractor - became a six year journey of environmental engineering, algorithm design, and behavioral insight. Palas concluded with a smile: even with a problem that “sounds easy-peasy,” agriculture always finds a way to surprise you. The work continues, he said, and every new season brings new lessons.

Content from this website can be adapted or shared, provided that Gurtam is credited as the author and a link to the original page is included. Unauthorized use without attribution is prohibited.