Why is AI flying while Robotics are flagging?

Agriculture has a genuine use case; fruit picking and weeding is back-breaking, hard, labor and the season is short. Robotics should be able to help, either in the form of mechanization with tractors and loaders or via robotic assistance…potentially using IoT connectivity, automation and remote control.

I had a word with a digital friend about this topic, to get some insider information from a genuine expert. The feedback was that virtuality is easy while reality is harder to build for. We know, IoT is connectivity built for real world applications. The real world environment is always changing, materials and parts are not standardized. There are few robot production lines constructing and repairing their counterparts for real world compatibility, yet.

Some of it seems too easy. they love e-mobility yet your car gets used for about 2 hours per day, the other 22 it sits in a shed waiting to work. Why not make add some solar panels to the roof panel and let it outside when the weather is right? Seems simple enough. Some of the big boys have been building humanoid or dog robots since the year zero.

Let’s backtrack slightly. What we are discussing here are the type of robots that appeared in TV and movies in the early 1970s, not mcuh more than a computer on some sort of mobility platfrom. The Star Wars model, made of metal with some sort of electronic processor. The Ripley from “Aliens” exo-skeleton concept is slow to gestate also, even though it seems reasonably feasible with modern materials, hydraulics, miniaturized motors and energy capture concepts.

Agriculture has many different crops, farm layouts, and techniques, making “one-size-fits-all” robots rare. Autonomous tractors, equipped with GPS and AI-driven systems, can be fine in flat, square fields of the American Midwest, but require significant upfront investment. The 100, 000 USD asking price for a tractor might be less of a head-scratcher for some, than configuring the options, but it prices most farmers out of the game.

A robotic harvester has to deal with rain, mud, bugs clogging sensors, and fruit growing in random places. Robots are limited by battery efficiency. A robot working in a warehouse (flat floors, fixed shelves) is much easier to develop than one picking strawberries in a field where lighting, plant growth, terrain and weather are subject to constant change. You see that controlled environments suit robots (as they are currently constructed). It is not an intersection of mechanical realities, but a simplification of the environment to fit the evolution of machines.

We discussed this topic previously when we referenced industrial robotics and their universal use in manufacturing processes. The key difference between agriculture and industry is that bulk human labor is a major asset; the work is repetitive, but it is not uniform. Until now it is still not scalable to invest in robotics (for most farms) when the price point of physical human labor is so much cheaper.

Ready to discuss your IoT project?

Let our engineers offer you quality technology consulting services.

Sign up for a free consultation.