Vermuë Arable Farming BV in Werkendam (N.-Br.) has acquired an AgBot T2 traction robot this year. Marijn Vermuë is testing how the robot can be integrated into his farm operations. He is eager to see what it delivers and what the real costs are. Marijn Vermuë aims to advance his arable farm, striving for maximum efficiency across his 160 hectares. His primary goal is yield maximization, but he also seeks to lower costs and reduce environmental impact through precision techniques. “Ideally,” he says, “we achieve both at the same time.”
Marijn joined his father’s farm in 2015. By 2020, he started using task maps to adjust the planting distance of potatoes based on soil density. In onions, herbicide dosage is varied according to soil weight, guided by soil scans and task maps. “Our fields don’t have a gradual soil transition from front to back; they’re patchy in terms of soil variation. I can’t adjust the dosage just by feel.” The foundation of these task maps comes from soil potential maps supplied by Van Iperen within the Precision Farming Plus project.
Van Iperen also provides planting recommendations based on germination tests. “Where the soil is lighter, I plant at wider spacing; where it is heavier, I plant more densely,” Vermuë explains. “I adjust between 20 and 30 cm planting distances. The theory suggests this should increase yield over a fixed planting distance, but I need test strips to confirm it.” In onions, variable herbicide dosing results in lower chemical use and reduced crop damage, leading to a higher and more uniform yield. Potatoes receive liquid fertilizer in the row at planting, improving efficiency compared to full-field application.
Pulsing nozzles for spot spraying is still a step too far for me. Too expensive.
Vermuë closely monitors developments in arable farming innovation. However, when it comes to precision agriculture, he is not exactly a pioneer. He follows the NPPL website and picks out what suits him and what doesn’t require extremely large investments. For example: “Pulsing nozzles on the sprayer to enable spot spraying and more precise variation is still a step too far for me. It would suddenly add €30,000 to the cost of my sprayer.”zijn.”
Now, Vermuë has decided to take a significant step forward. Under the flag of the National Precision Agriculture Testbed (NPPL), he will be working with a field robot throughout the entire season. Starting with mulching cover crops and preparing the land, and ending with soil cultivation after harvest. Vermuë also plans to use the robot for sowing cover crops.
This involves the tractor robot or multi-purpose AgBot T2 from manufacturer AgXeed. The project is part of the NPPL-R initiative, which will be rolled out in 2025. Various field robots will be deployed with arable farmers across the Netherlands. The AgBot is an electric-powered tracked tractor with wide tracks. A Deutz diesel engine powers the generator. Vermuë’s existing implements can be attached via the three-point hitch. The weight is 6 tons with an output of 156 horsepower. Vermuë mentions that, especially for spring work, it is advantageous that the AgBot is on the lighter side and operates on 76-centimeter wide tracks, which reduces the risk of soil compaction.
It is estimated that around 400 hectares of work will await at Vermuë Akkerbouw’s locations in Werkendam and Dreumel next season. Vermuë is considering using the AgBot for contract work as well. It will mainly involve simpler tasks: pre-tilling, soil loosening, cultivating, preparing seedbeds, and similar tasks.
A big question remains how the obstacle detection system will respond to flying soil and root weeds when the Kvik-up is attached to the robot. The tines of this root weed control device work through the soil, throwing soil and root fragments into the air. Will the sensors identify the airborne clumps and weeds as obstacles or not? And how can this be adjusted without compromising safety? Vermuë: “What definitely won’t work is keeping the material low with a flap. The trick of the Kvik-up is that the heavier soil particles fall back down faster than the root fragments, allowing them to land on the surface and dry out.”
FutureFarming is a media partner of Boerderij and tested the AgBot from AgXeed last year in Australia. The robot operated for 500 working hours in its first year.
There are other issues at hand that need to be addressed. What Vermuë is concerned with is determining what such an AgBot can mean for his business in the coming period. “I can still hire people, but sooner or later that will become more difficult. Moreover, this robot can work around the clock. It doesn’t need to visit the dentist, go to bed, or have social obligations.”
It will need to be clear next year how much work is required to manage the machine, and therefore how many man-hours it will save Vermuë. And, of course, important: Vermuë wants to use the trial period with the tractor robot to gain insight into its business case. The purchase price of the machine is approximately €325,000.