The July 2022 issue of IEEE Spectrum is here!

Close bar

Bosch's Giant Robot Can Punch Weeds to Death

A modular agricultural robot from Bosch startup Deepfield Robotics deals with weeds the old fashioned way: violently

2 min read
Bosch's Giant Robot Can Punch Weeds to Death
Photo: Deepfield Robotics

At IROS last month, researchers from a Bosch startup called Deepfield Robotics presented a paper on “Vision-Based High-Speed Manipulation for Robotic Ultra-Precise Weed Control,” which has like four distinct exciting-sounding phrases in it. We wanted to write about it immediately, but Deepfield asked us to hold off a bit until their fancy new website went live, which it now has. This means that we can show you video of their enormous agricultural robot that can autonomously detect and physically obliterate individual weeds in a tenth of a second.

imgImage: Deepfield Robotics

Given the scale of farming today, treating weeds chemically is really the only practical way for humans to keep them under control, because you can use tractors or airplanes to cover large areas in a short amount of time. But all of those necessarily deadly (to weeds) chemicals then get on the plants we don’t want to kill (because we want to eat them), as well as getting washed into the soil. The most organic and eco-friendly way of dealing with weeds is the old-fashioned way: physically removing them. “Physical removal” can mean pulling weeds out completely, but that involves both grasping the weed and doing something with it. A better solution is to just smash it way down into the ground, which is faster, easier, and something a robot can do excellently:

The stamping tool is 1 centimeter wide, and it drives weeds about 3 cm into the soil. It’s designed to detect (through leaf shape) and destroy small weeds that have just sprouted, although for larger weeds, it can hammer them multiple times in a row with a cycle time of under 100 ms. Testing on a real carrot crop, which has carrots spaced about 2 cm apart and an average of 20 weeds per meter growing very close to the carrots themselves, the robot had no trouble at all. The maximum capability of the system is about 1.75 weeds per second at a speed of 3.7 cm/s and a weed density of 43 weeds per meter, but at lower weed densities, the speed can be cranked up to 9 cm/s.

The robot in the video looks very much like a research project, but it’s intended to be a module that can be used on Deepfield’s “adaptable multi-purpose robotic platform,” BoniRob.


Image: Deepfield Robotics

BoniRob can navigate itself, adapting to many different field configurations. Its modular payload bay can handle up to 150 kilograms of stuff, and an onboard generator lets it run autonomously for 24 hours without needing to refuel. It’s powered by ROS, and Deepfield even suggests that you could use it for crazy stuff like drone launches or even swarms of mini-BoniRobs, because why not [right].

The idea here is that farms could buy one BoniRob, and then buy or rent whatever modules they happened to need when they need them, without having to invest in many single-task robots. With Deepfield currently conducting extensive real-world tests with BoniRob operating autonomously on farms, we’re guessing that it’ll be available in the very near future.

“Vision-Based High-Speed Manipulation for Robotic Ultra-Precise Weed Control by Andreas Michaels,” by Andreas Michaels, Sebastian Haug, and Amos Albert from Bosch, was presented last month at IROS 2015 in Hamburg, Germany.

[ Deepfield Robotics ]

The Conversation (0)

How the U.S. Army Is Turning Robots Into Team Players

Engineers battle the limits of deep learning for battlefield bots

11 min read
Robot with threads near a fallen branch

RoMan, the Army Research Laboratory's robotic manipulator, considers the best way to grasp and move a tree branch at the Adelphi Laboratory Center, in Maryland.

Evan Ackerman

“I should probably not be standing this close," I think to myself, as the robot slowly approaches a large tree branch on the floor in front of me. It's not the size of the branch that makes me nervous—it's that the robot is operating autonomously, and that while I know what it's supposed to do, I'm not entirely sure what it will do. If everything works the way the roboticists at the U.S. Army Research Laboratory (ARL) in Adelphi, Md., expect, the robot will identify the branch, grasp it, and drag it out of the way. These folks know what they're doing, but I've spent enough time around robots that I take a small step backwards anyway.

This article is part of our special report on AI, “The Great AI Reckoning.”

The robot, named RoMan, for Robotic Manipulator, is about the size of a large lawn mower, with a tracked base that helps it handle most kinds of terrain. At the front, it has a squat torso equipped with cameras and depth sensors, as well as a pair of arms that were harvested from a prototype disaster-response robot originally developed at NASA's Jet Propulsion Laboratory for a DARPA robotics competition. RoMan's job today is roadway clearing, a multistep task that ARL wants the robot to complete as autonomously as possible. Instead of instructing the robot to grasp specific objects in specific ways and move them to specific places, the operators tell RoMan to "go clear a path." It's then up to the robot to make all the decisions necessary to achieve that objective.

Keep Reading ↓Show less