How do farms connect machines without cell coverage?

Precision agriculture runs on data. USDA’s National Institute of Food and Agriculture describes the pattern: data are collected at high spatial or temporal resolution, analysed, and turned into treatments applied by equipment capable of precise control. A combine’s yield monitor records grain flow with satellite positioning, a soil map drives a variable-rate prescription, and the applicator records what it actually applied.

Guidance technology in particular has spread. USDA’s Economic Research Service (ERS) reported in 2023 that automated guidance was applied on well over 50 percent of the acreage planted to corn, cotton, rice, sorghum, soybeans and winter wheat. Each of those machines produces data that has to go somewhere.

Between tractor and implement, data stays on the machinery. The Agricultural Industry Electronics Foundation (AEF) explains that the ISO 11783 standard, known as ISOBUS, defines communication between agricultural machinery, mainly tractors and implements, and data transfer between those machines and farm software. The aim is plug and play for any combination of tractor, terminal and implement.

In practice, the AEF notes, how well components work together depends on the functions they all support, and the standard leaves room for interpretation. The AEF writes guidelines that narrow that room, which it says gives clarity about the compatibility of ISOBUS products, and it runs a conformance test.

Getting data from the field to the office is where coverage matters. ERS defines telematics as the wireless transfer of data between farm equipment, connected devices and the cloud, either through local area networks such as Bluetooth or through cellular systems. Yield data are recorded on logging devices in the machine and pushed to cloud systems in near real time over cellular links when they are available. Prescription maps travel the other way, down to the applicator.

The same report describes the gaps:

• In crop-producing regions, wireless connectivity may be weak or nonexistent, especially for fields away from interstate highways or towns.
• Internet providers typically favour download speed, while field operations need to upload large files.
• A lack of broadband access has been identified as a major barrier to adopting digital farming tools.
• Whether newer networks will help depends on whether 5G reaches the same areas as 3G and 4G did, and whether satellite internet offers low enough latency.

Without telematics, data moves on flash media carried from the machine to the office computer. ERS notes that this is a simple task, but that lost cards have been a barrier to making good use of the data, and that telematics improves on it by automating this manual step. It is a form of data muling: the record travels physically with a person or vehicle until it reaches a connection.

The result is a mix: telematics where the signal reaches, and manual transfer or delayed upload where it does not.

Device-to-device software makes the carrying automatic. A phone in the cab, a tablet on the implement, or a sensor gateway at the field edge can exchange records directly with nearby devices, over local links of the kind ERS lists, without waiting for a tower. Records then wait in an outbox and move on, hop by hop or with whoever drives back to the yard, until one device reaches coverage and delivers them to the farm management system.

Offline Protocol’s mesh SDK is one example for the phone side: nearby phones exchange MLS-encrypted messages over Bluetooth LE with no tower, unsent messages stay in an outbox, and the farm’s backend stays the system of record.

Three questions decide whether this helps a given farm:

• What must arrive, and how soon? A prescription map needed before spraying differs from a yield log that can reach the office at the end of the day.
• Who carries the data out? A truck, a phone or a fixed gateway near the farmhouse each give a different delay.
• When is a record finished? The local copy should stay until the farm system confirms it accepted the record, so nothing is lost if a device breaks on the way. Handing a record to another phone in the field is not the same as the farm system accepting it, and the app should show which stage each record has reached.