The challenges of working with autonomous platforms | Field Trials E8 PART 3

The more work invested into mapping the field, the more likely autonomous machines will operate without hiccups. That’s the idea, anyway.

According to Saskatchewan farmer Dustin Burns, making AI-based systems work efficiently takes a lot of time, as well as a change in one’s overall production mindset.

Part three of Burns’ Field Trials discussion covers the challenges associated with practical field operations of autonomous platforms such as Raven’s OMNiPOWER – a machine which Burns has employed and field-tested for four years. This includes the need for significantly more up-front planning, and an understanding that autonomous platforms may have to work much slower than initially desired.

“If we can have autonomous units that we send on a mission and take their time, do a better job, apply fewer chemicals and do as good a job…there’s some real gains there,” says Burns. “If we’re processing information on the fly, it might require in the short term to slow down our application speeds, but in the end, it may save you money or comply with regulations in the future.”

“We’ve got fixed field boundaries, but they’re still working on the AI piece where they can bring in perception and understand basically what an operator does. If I’m going down the field and I see an obstacle, whether it’s a big rock or a person, somebody drive into the field, right now how do you do avoid an accident? But on the other hand, I think, well, with autosteer on our conventional equipment and you get distracted doing something else, is the risk any more or less?”