Robotics

SwagBot autonomous agricultural robot to hit the farmers market in 2020

View 2 Images
The SwagBot autonomous farm robot has received enough funding to commercialize it by 2020
Agerris/University of Sydney
AI enables autonomous weed recognition and treatment
Agerris/University of Sydney
The SwagBot autonomous farm robot has received enough funding to commercialize it by 2020
Agerris/University of Sydney

Back in 2016, we wrote about the Agerris SwagBot farming robot and its abilities in crop inspection, weed removal and yield counting. Now, having raised US$4.6 million for commercialization, the company is looking to get an AI version of the robot onto the market within a year.

Where the 2016 version was remote-controlled, the SwagBot can now work as an autonomous unit, with the capability to identify and eradicate weeds, while monitoring pastures and crops. It'll soon have the ability to autonomously herd cattle built in as well, so Spot the farm dog might find himself out of a job unless he can learn to monitor and wirelessly report on the herd.

Where SwagBot is designed to focus on large-scale crop and livestock operations, Agerris is also working on a second product called Digital Farmhand for smaller farms.

Digital Farmhand will automate tasks like non-chemical weed removal, intelligent crop spraying and yield estimation. It'll work best for row and tree croppers, and Agerris is hoping it'll be price-competitive enough to be useful in developing countries.

The company is targeting a commercial release within 12 months, starting in its home country of Australia but expanding globally after that. There's still some testing and development work to be done, however.

Whatever the commercial fate of the SwagBot and Digital Farmhand, the agricultural robotics market seems set to explode in the next decade or so, with robust, autonomous mobile platforms developing alongside deep learning and artificial intelligence to a point where their utility and affordability will be very compelling.

See a demonstration of SwagBot in the video below.

Source: Agerris

  • Facebook
  • Twitter
  • Flipboard
  • LinkedIn
2 comments
DaveWesely
As someone raised on a farm, and a robotic lawn mower owner, it is depressing to see how far off the mark our first attempts with robotics in farming have been. First off, the money isn't in weeding pasture land. It is in row crops. The vast majority of row crops are large plants - corn (10ft), soybeans (4ft). It makes no sense to straddle a row once the plants are more than a couple of feet high. That is why GMO crops are so popular with farmers. They can broadcast spray herbicides to kill weeds without killing the crop. Agribusiness can sell a lot of herbicide that way. The next thing is mobility. Wheels don't work. The last thing a farmer wants to do is spend their day rescuing robots from ditches and mud. Tracked vehicles may not be as energy efficient, but they don't get stuck. So once you have a tracked robot with the proper electric PTO and hitch connections that can navigate down rows and recognize crops vs. weeds, then maybe you have a game changer. You would be able to tend the field throughout the growing season, instead of just the spring and fall. Your herbicide and seed bills would drop dramatically. Herbicide resistant GMOs would not be necessary, since you could spot treat individual weeds (we can program computers to recognize faces, we can do the same to recognize weeds). Robotics would also allow the farmer to do stuff like plant green manure before an existing crop is harvested, or selectively plant and harvest crops based on field conditions.
Mark Pettit
I wonder if this could be used to pick plastic off beaches?