Field robot or retrofit autonomy? Dutch grower tests both

A Dutch arable grower has run two very different autonomy systems back to back: a full season in 2025 with an AgXeed AgBot 5.115T2 field robot, then a week in mid-October 2026 with a New Holland T6.180 fitted with GPX Solutions' iQuus retrofit kit. Marijn Vermuë, who farms in Werkendam, North Brabant, concluded both systems are mature enough to justify an investment — but they deliver different advantages, and the right choice depends on the farm.

What did he test?

The AgBot 5.115T2 is a cabless, autonomous tracked tractor that Vermuë worked with throughout the 2025 season, including combinations such as a Kverneland 3400 B plough. The iQuus kit turns a conventional tractor into an autonomous one while keeping it drivable by an operator for tasks such as transport. During the October trial, Vermuë focused on operation, ease of use and versatility rather than tracks versus wheels or overall weight.

He is clear this was not a competition. "Comparing autonomy systems is not simply a matter of using a tape measure, weighing scales or a points system," he notes. The weight each factor carries depends on the individual farm and its operations.

How easy is the iQuus system to run?

Operating the iQuus-equipped tractor is "very straightforward", says Vermuë. "The manual takes you through the programme step by step. Foolproof. You cannot go wrong."

The numbers back that up. On Tuesday morning, 13 October, it took less than an hour to get the New Holland cultivating a field after potatoes — its first autonomous job. In the next field, the combination was running in under 10 minutes.

Set-up time matters more than many buyers assume, he argues. "If 10 people end up working with an autonomous tractor over the course of a season, I think it needs to be quick and easy. I do not have unlimited time. What I do not want are things that quietly and unintentionally eat up minutes all over the place."

Route planning shows little difference between the brands. "The iQuus route planning is certainly as practical as AgXeed's. They are slightly different, just as Apple's iOS and Android differ. You need a little time to get used to them."

How do the safety philosophies differ?

The systems take opposite approaches to field boundaries — the geofence, the virtual line the machine will not normally cross. GPX Solutions treats the boundary as advisory: the operator decides whether to work beyond it. AgXeed treats it as absolute.

iQuus requires only the implement's working width to be entered, saving time at set-up. But its field boundary is not absolute — implements can swing beyond it. Vermuë: "That is often not a problem and can even be efficient alongside a ditch or farm track, but you need to be careful around obstacles. With the AgBot, the geofence is 'sacrosanct'." He compares the two safety philosophies to Apple's closed ecosystem versus Android's openness.

Each approach has a cost. iQuus lets the plough turn over above the ditch, gaining workable area, but a badly planned route could put the tractor in the ditch. AgXeed never allows an implement into the neighbour's hedge, but it will also never let one extend over the ditch. "As farmers, we will always look to work right up to those boundaries," says Vermuë.

During start-up, iQuus issues clear warnings if a route extends beyond the field or passes too close to obstacles, stating that it may be dangerous to drive the route autonomously. Vermuë: "You can then decide whether to adjust the route or go ahead anyway, at your own risk, of course." Joanne de Ruiter of GPX Solutions confirms the company deliberately avoids a fully restrictive geofence: "We could, but then customers would say: 'We do not want it.'" Customers make that call themselves, she says, just as they would behind the wheel.

What are the practical gaps?

Transport differs sharply. The AgBot needs a low-loader to move between farms; the iQuus-equipped tractor drives itself to the field, but the operator then needs a lift home. Vermuë sees little overall time difference and suggests a practical fix: "I would rather not stand around for 10 minutes waiting to be picked up. Another option is to take a bicycle or fatbike along on the tractor."

One functional gap remains: iQuus currently cannot work headlands autonomously, so an operator must drive those passes. "I understand it will not be long before iQuus can do this too, but AgXeed is not standing still in the meantime either," Vermuë says.

What does it cost?

Vermuë wants autonomy across two tractors — a heavier one for cultivation and a lighter one for crop care. GPX Solutions broke down the pricing for that scenario:

• A single iQuus system costs approximately €50,000.
• Half of that — €25,000 — covers components permanently installed on the tractor.
• The other €25,000 covers transferable components.
• Preparing each additional tractor costs €25,000, bringing two tractors to €75,000 in total.

"What appeals to me is that you could perform heavy tasks such as deep spading autonomously for a relatively modest additional investment," says Vermuë.

For growers weighing the same decision, the watch points are clear: iQuus headland autonomy is expected soon, AgXeed continues developing, and transferability between tractors may decide which system fits mixed machinery fleets.