For the past few years on the UC Santa Cruz farm, a team of students known as the Electrified Slugs has been iterating on software to control a small electric tractor, making it capable of self-driving along plant lines and pulling weeds.
This work led the Electrified Slugs to turn their attention to another challenge: the critical issue of worker safety within human-robot interaction. After more than a year of development, the team came up with an emergency stop mechanism for the tractor and a method to control the tractor via a smartphone. Their efforts were recognized with the “Excellence in Safety” award at the second-annual Farm Robotics Challenge.
The team included then-Robotics Engineering undergraduates and now Electrical and Computer Engineering (ECE) master’s students Katherine Rogacheva and Oliver Fuchs, Computer Science and Engineering undergraduate student Milos Suvakovic, and Robotics Engineering undergraduate student and Frontiers Fellow Sam Leveau. Professor of ECE Dejan Milutinovic and Executive Director of the UC Santa Cruz Center for Agroecology Darryl Wong were team advisors.
“What has inspired me most about this team is their understanding that robotics, autonomy, and AI have the potential to improve practices and sustainability, but that there needs to also be a human integrated into the work to support continued learning and adjustments,” Wong said. “This tech is not about taking people out of the fields, but about making it safer, easier, and more humane to do the work.”
The Farm Robotics Challenge was organized by University of California Agriculture and Natural Resources and the AI Institute for Next Generation Food Systems, with support from technology partner farm-ng. Farm-ng, based locally in Watsonville, produces the Amiga, a modular robotic platform for many agricultural uses from small, organic farms to commercial growers. The UC Santa Cruz team was among five winning teams from colleges and universities across the country.
When announcing the Electrified Slugs’ award, the judges praised their solution’s connectedness to industry needs, the way the team challenged their own skills, and the delivery of a slick design.
"You are forward thinkers who continue to improve and advance what is possible for these Amigas," the judges’ comments concluded.
Since the first Farm Robotics Challenge in 2023, at which they won the Small Farm prize, the UCSC team has been experimenting with software and hardware tools to add to the Amiga to address challenges on small, organic farms.
“At UCSC with the Center for Agroecology, we’re very geared toward organic farming, which is awesome. If there’s any way we can improve their process or help make farming better or less labor intensive for the workers, that would be a good goal for us,” Rogacheva said.
In their first year of work, the team developed software to autonomously weed plant lines, and experimented with using the system for seeding and tilling, with the latter efforts led by agroecology student and former Electrified Slug team member Mauricio Chavez.
Weeding, seeding, and tilling require close interaction between the farmer and the robot, bringing worker safety issues top of mind. In the agricultural industry, tractor accidents are the leading cause of fatalities on U.S. farms. The accidents can also lead to workers incurring long-term disabilities and farms to go out of business.
The group of students spent several months developing new mechanisms to improve the safety of the Amiga, landing on two main solutions: an improved emergency stop and a remote navigation joystick. Both of their solutions are controlled from an app on a mobile device operated by the farmer.
“It’s a very sleek design, and it was very fun to make,” Milutinovic said. “It’s also very high tech — our students actually get to work with all aspects of autonomy, remote control interaction, mobile devices, communication, and imaging. It’s a useful and interesting combination of knowledge.”
“I’m interested in the connection between something very new like robotics, into something very historical, like farming — what’s the best way to get an easy transition,” Leveau said. “What we wanted to do was develop a user-friendly app, that lets a farmer use a piece of technology that's new to their farm without being overwhelmed, and with the right to repair the machine themselves.”
The software preloaded on the Amiga includes a stop button which shuts off power to the machine. On flat ground, this is effective, but on a slope, the Amiga can still roll and move. The team’s solution introduces a non-mechanical stop, sending a command that forces the Amiga to stay at zero velocity.
To further improve safety, the team built out a method of controlling the movement of the tractor with a joystick on the mobile app, an alternative to the standard joystick which is physically connected via a cable to the Amiga and could interfere with the wheels. Within the same app, they also added a feature that enables the farmer to remotely monitor the tractor’s camera feed and control its movements. The team also continued to develop their vision-based plant detection software, an effort led by Suvakovic.
The team received a $2,500 prize with their award. The student team received a seed award from CITRIS at UC Santa Cruz to fund their efforts.