Resisting robots or embracing automation?

What role should robots play in agriculture?

The use of robots in the workplace continues to be a contentious issue as concerns over worker displacement and disempowerment overshadow potential gains in efficiency and safety.

Agriculture is a little different, as the sector faces a combination of labour shortages and aging demographics. However robots in agriculture is contentious for a range of other reasons, that can and should inform how robots are used in the broader economy.

The automation debate is never just about technology, or business, but rather the relationship between robots and humans, and the way in which automation impacts the farm, and the extent to which farmers have influence or agency.

The issue of automation is really an issue of control and configuration. Robots are not rigid, but rather programmable. The larger question is who does the programming.

As an industry, agriculture is diverse, but there is an underlying ethos of innovation, resilience, and do-it-yourself (DIY). The DIY ethic is particularly important, as agriculture can be incredibly time sensitive. Nature waits for no one, and farmers often don’t have time to fuss with equipment or wait on repairs.

As a result the biggest resistance or opposition to automation comes from the current constraints or obstacles against user control. For example the right to repair is a growing concern, as many high tech agricultural tools and equipment are effective black boxes, controlled by their manufacture, and not open to modification or repair by the user or farmer.

This is a big problem, and one that has to be addressed, if automation is to make further headway into the broader agricultural sector.

For example, an ongoing UK pilot project examining the automation of agriculture is the Hands Free Hectare project run by Adams University and Precision Decisions:

While this project is largely focused on and derived from autosteer technology, it provides glimpses into how automation will be employed and adapted for farming.

One of the advantages of autosteer is that it can be added to older tractors and does not require full integration into the heavy machinery. This enables the operator to be able to repair the machine without having to deal with proprietary software or interdependent systems.

However autosteer is just a tiny example of automation. A powerful one, given the size of the machines involved, and the tasks they can perform, but just the tip of the iceberg of what is possible. The danger is using autosteer as a kind of slippery slope to seduce farmers into larger applications that result in the loss of control of their equipment or the application.

This is why the right to repair and open source in general is so important. Automation has to benefit all parties involved, and not just the vendor, manufacturer, or supplier.

It may be one of the reasons that other agricultural robots have also faced some push back:

Could a robot actually take on the job of a sheepdog?

The footage comes from New Zealand firm Rocos, which announced a partnership this week with Boston Dynamics, maker of the four-legged Spot robot that stars in the video (and many others). Rocos makes software to control robots remotely, and the video demonstrates one potential use-case: agriculture.

“Equipped with payloads like heat, LIDAR, gas and high resolution camera sensors, Spot navigates rugged environments to capture data in real time,” says the company in a blog post. “In agriculture, farmers can access information such as more accurate and up-to-date yield estimates. This provides access to a new category of automation, and a safer, more efficient business.”

Rocos is a company that demonstrates the need for an ecosystem to enable the usefulness of this technology. In this case by partnering with one of the industry heavyweights, Boston Dynamics, Rocos has a robust robot, that they’ve created a user interface for users to control.

This is one possible scenario as to how robots can migrate into farming, by making it possible for farmers to program and control the robot directly. However it also ignores the larger body of knowledge and evolution that agriculture has fostered over many centuries.

The robot might be an amazing tool for lots of things but it is worthless and unwanted as a sheepdog,” Rebanks told The Verge. “No one who works with sheep needs or wants this — it is a fantasy.”

Rebanks says robots simply don’t have the motor skills or the intelligence needed for such demanding work, and they likely won’t for a long time to come.

“Moving sheep isn’t just being behind them, it is about doing whatever the controller asks, and sometimes what needs doing based on [the dog’s] own intelligence beyond the handlers control,” he says. “A shift to the left or right of a few inches can turn the sheep, and a great dog can judge their characters and how much to do or not do.”

This relationship between sheep and dog — the dynamic of two intelligent beings — is vital, says Rebanks, and it’s rooted in the evolutionary history of predator and prey.

The necessary ecosystem to make this technology work has to involve more than just an interface, but an ability to respond and work with existing expertise. In this case, that means not just the farmer or shepherd interacting with the robot, but the sheepdog as well.

This matches our larger vision of automation, that it is not worker vs machine, but rather workers plus machine. It’s about augmenting and expanding the role of the human (or dog) rather than replacing them.

Indeed the real potential of robots and automation is not just in expanding operations, or using labour more efficiently, but in enabling better agricultural practices. This is the real promise of precision farming. Not only can costs be reduced, but so too can the use of pesticides, herbicides, and other inputs.

In a future issue we’ll talk about the role of robotics in fighting weeds, pests, and profile a company that wants to use robots to enable regenerative agriculture and large scale carbon farming.