At recent farm shows, John Deere has captured the spotlight with pictures and descriptions of an autonomous tractor. And just like six years ago, when Case IH debuted a concept model of an autonomous tractor, farmers were intrigued.
I am not a skeptic of self-driving vehicles but have some thoughts about the adoption of this technology. One pertinent example is mine haul trucks. The Komatsu autonomous haul truck announced in 2016 was one of the first companies. Six years later there are about 800 autonomous haul trucks in operation globally, over half in Australia. They represent about 1.5% of all such vehicles.
Large open-pit mines are a logical early application for this technology – tightly controlled routes and traffic. However, this technology is a staggering leap, and the mining industry staggering to make that leap, even in this straightforward environment. This is often typical of technology adoption. This simplified graph illustrates a rough summary of what researchers have discovered since beginning to study this phenomenon. Not only is it really slow, even for gee-whiz inventions, there is a crucial chasm where earlier adopters struggle to convince the large majority. This is also a crucial point in a manufacturer’s business plan – the profits are on the right side of the chasm.
Autonomous farm machinery faces another nearly unique constraint. These are plat maps of my township a century apart. The land is divided into virtually the same number of fields - about 250. I don’t see any pathway to massive contiguous tracts where autonomous machinery can pay for itself. Our land ownership pattern will impede deploying such machines far more than cost or expertise. Just moving from field to field is an ongoing constraint on machine size, maneuverability, and operator expertise. This technology will show up sooner in places like Brazil or Ukraine, where field size is measures in hundreds if not thousands of hectares. An alternative strategy for our Midwestern geography might be multiple smaller and cheaper autonomous units that fit fields and roads better.
Autonomous technology may be adopted reasonably soon, but I think almost exclusively on those farms who post the combine-fleet drone videos we’ve all gotten tired of.
