For farmer and inventor Brian Tischler, the question isn’t whether autonomous tractors are cool, possible, or useful — it’s how to overcome the liability risk if something goes wrong.
Tischler told farmers attending CropConnect in Winnipeg recently it’s possible to build your own self-driving vehicle for around $1,000 using open-source software. He’s done it.
That’s made the Manville, Alberta farmer a popular presenter on the winter meeting circuit over the past couple of years. He and his wife Arvel work with two business partners to operate Genetic Potential Ltd., a collaborative mixed farm raising cattle and multiple crops.
Why it matters: Autonomous driving farm equipment is seen as one option for addressing the human resources gap on Canadian farms — but taking human hands off the controls is risky.
Tischler does hardware development and programming as a hobby, making programs such as AgOpen GPS freely available through open source hubs as a way to share information and give back to the farming community, he said.
Tischler cautioned that being able to do something with technology isn’t the same as saying you should do it.
“Is it cool? Yes. Is it possible? Yes. Question is, should we? That’s the real question,” he told his audience of about 200 farmers.
There are lots of compelling reasons why autonomous driving implements might be helpful in agriculture, including more efficient use of human resources, better precision and even safety.
The sensors and guidance systems in use today are much less likely to fail than humans who can tire, become distracted and make poor judgement calls.
However, society has a much greater tolerance for human error than it does for equipment failure. As such, a human accidentally driving over someone with a tractor is seen differently than if it was an autonomous tractor.
In his view, that’s the main reason why major farm equipment manufacturers have had the capability for autonomous machines since the 1990s, but are reluctant to bring it to market.
“Your insurance company is going to puke,” he warned farmers who might be considering trying this at home. “I’ve talked to my insurance company, I’ve spoken to a lawyer. It’s obvious. You’re responsible. The manufacturer is responsible. If something goes wrong, everyone gets sued and it will destroy you. That’s it.
Tischler cited airbags made by the Japanese company Takata. While airbags have saved hundreds of thousands of lives worldwide, a series of incidents in which people were killed by defective airbags manufactured at one of the company’s plants prompted global recalls and put the company out of business.
He said the safety mechanisms can’t just be about shutting down software if something goes wrong, they must be able to address a scenario in which something as simple as the clutch fails to engage.
“This is definitely not a case of replacing a driver with a remote control. It has to be able to think and drive its surroundings electronically and mathematically and react and keep safe and keep out of trouble,” he said. “It’s got to work and it’s got to be proven to be working.”
He noted self-driving cars have the potential to reduce the number of people killed in traffic accidents, which is the equivalent today of a 737 crashing every hour.
“But if they take one life because it is self driving that’s the end of that company, which is irony at its best,” he said. “So it is the same in agriculture.”
“The technology is there and the software is there, the question is — should we? And how much trouble are we going to get into?”