One field, two harvests: crops and clean electricity. Agrivoltaics promises to transform farmland into a powerful climate solution, while autonomous robots reduce labour and optimise operations. Experts say the technology is ready, but regulation is not.
by Cornelia Trefflich
Producing fruit and electricity at once: when robots meet agri-PV
Picture a pear orchard in Randwijk, the Netherlands: just endless rows of "Conference" pear trees.
Between them rolls a four-wheeled vehicle resembling a lawnmower, fitted with a camera about a metre and a half above the ground. It goes about its task undeterred, driving systematically along the rows of trees and photographing them as it passes.
No driver, no engine noise, no fuel: yet it finds its own way, including back to the charging station at the edge of the field. A glance upward reveals what sets this orchard apart:
rows of solar panels mounted high above the trees, high enough to leave the pear trees below room to grow.
"You do need your plants to grow, so most of the constructions are either elevated as with the pear trees that we have now or spaced wider apart than conventional solar parks [...]," explains Hellen Elissen, senior researcher in agrivoltaics and project leader at
Wageningen University & Research (WUR).
A single plot of land thus produces two things at once: fruit and electricity. In addition, an unmanned robot handles the laborious monitoring work that previously required an entire team.
What sounds like a vision of the future is being tested here under real-world conditions. For three years, Wageningen University & Research (WUR) has been
studying how yield, fruit quality and soil conditions change beneath the solar panels. This test field is part of the
EU-funded research project TALOS, which investigates
how robots could assist farmers in the future. But what exactly is behind this project and what benefit does it offer to farmers in Europe?
Twice the purpose, five times the potential: the agrivoltaics paradox
In technical jargon,
this novel technology is known as agri-photovoltaics, or agri-PV for short. It refers to the simultaneous use of a piece of land to grow crops and generate solar power.
"The main aim of this concept is dual land use: on the same plot you are growing the crops and also producing electricity," explains Jiří Bím, head of the agrivoltaics section at the
Czech Solar Association. The main advantage lies in optimal land use:
a plot no longer has to serve exclusively as farmland or as a power plant but simply delivers both at once, as Bím explains:
"You grow 60% of the potatoes, and 70% of the electricity. In total, you have 130% efficiency of using this land. If you separate it, you need 1.3 hectares to get the same yield." This holds enormous potential:
according to a European Commission JRC study,
installing agri-PV on just one percent of the EU's agricultural land would yield around 944 gigawatts of capacity, more than double the
roughly 406 gigawatts (DC) of PV capacity installed across the EU by the end of 2025.
The panels also shield the crops underneath them from heat, heavy rain, and hail, and they slow down evaporation. Elissen gives an example:
"With raspberries, they are already covered in summer with plastic foils to protect them on the hottest days. If you replace those with solar panels, you have a win-win situation. You're producing electricity, and you're avoiding the use of plastics."
Agri-PV is also gaining importance within the EU for several reasons.
The EU plans to increase its solar capacity to almost 600 gigawatts (AC) by 2030, equivalent to roughly 750 gigawatts in DC terms, the unit commonly used by the solar industry. At the same time,
land is becoming scarcer and agriculture is increasingly exposed to the vagaries of climate change. The EU has recognised this and presented agri-PV as one possible solution.
The technology is furthest advanced where it builds on established methods. Unlike in the Netherlands, installations in France and Italy shade vineyards and orchards. In Germany and Austria, for instance, the panels are arranged in rows between crops so that tractors can drive through to harvest potatoes or cereals. Expert Bím confirms that this approach can bring benefits:
"We were in the Perpignan project, and the farmer told us he can produce 20 percent more wine, in better quality, with the agrivoltaics site."
When Robots and AI go to work in the field
Agri-PV is already in use at a number of sites, but a field that produces both crops and electricity also needs to be monitored twice over – once for the crops, once for the panels. This is precisely
where the EU-funded research project TALOS comes in. The project investigates
how robots and artificial intelligence can automate the maintenance and monitoring of solar installations: inspecting, cleaning and keeping vegetation in check.
Elissen sees this potential as well: “[...] once this technology is established, farmers can have the benefits of this autonomous technology added to it. So you basically can combine it with precision agriculture, which can lead to decreasing costs, it can lead to less labour, because pear and apple production involves a lot of seasonal labour, which can also be very costly. So that's the whole idea of the project, that you actually make work safer and that the farmer has to invest less in labour and the maintenance of his project."
To make the machines work together, TALOS is developing a software platform that coordinates the joint deployment of various machines, such as drones and mowing robots. Humans, however, are still needed for guidance and decision-making.
In practice,
the systems are being tested in three different settings: on land, on water (floating solar installations), and on agricultural solar sites. After three years, at least five finished robotic and AI solutions are expected to be in place. The
targets are specific: a
10 percent increase in solar park output, a 35 percent reduction in water consumption, more than 450 fewer tonnes of CO2 per year, and a 90 percent reduction in risk to workers. For agri-PV specifically, there is an additional target: a 90 percent reduction in manual effort for harvest monitoring. But how exactly is this device meant to achieve such an ambitious goal?
The "Husky": opportunities and limitations
Back in Randwijk, the agricultural robot mentioned at the outset makes its rounds: just over a metre long, 60 centimetres wide, weighing 50 kilograms, and going by the name "Husky." This unmanned ground vehicle (UGV) operates completely autonomously as
it moves through the pear orchard, finding its way via satellite signal and LiDAR, a sensor that emits invisible infrared light and uses their reflections to build a three-dimensional image of its surroundings. Mounted at a height of one and a half metres are
cameras that scan the rows of trees on either side. What they see is interpreted by an AI trained on thousands of images:
"Viewer is a camera system that can be mounted on tractors or any machine operating inside the farm, and it's the virtual eyes of the farmer. This system can take thousands of images, process them very fast, and give back to the far for each tree: how much fruit is on each tree, if there are certain diseases. It helps farmers monitor and observe their field at superhuman speed at scale”, explains Nikos Mylonas, co-founder and CEO of
EdenCore, the Greek company that makes and sells the Viewer.
Compared with laborious manual work, the efficiency gain is enormous, says Nicolas Congouleris, a robotics engineer at the
Centre for Research and Technology Hellas (CERTH) which developed the robot's energy management system.
But the
test field's real-world conditions in the Netherlands put the autonomous robot through a severe test, as Congouleris explains:
"We trained it in Athens, and navigation with just the LiDAR worked perfectly. And then it arrived in the orchard, and it just completely got lost. Because the environment is so homogeneous, the robot could not localise itself. It would enter the orchard, localise accurately for an inconsistent amount of distance – sometimes half a row, sometimes three – then suddenly think it is in a completely different row."
The researchers solved this problem by combining LiDAR sensors within the rows with GPS corrections at the vehicle's turning points; the Husky thus only corrects its position by satellite once it emerges from the rows of fruit trees. The solar panels also affect the robot's results: they sometimes block the GPS signal, complicating orientation, and their shadows make it harder for the AI to count fruit accurately. Flexibility was also required for battery charging, since the test field is used for other experiments as well and no fixed solar charging station could be installed there. As a result, the robot now has to recharge at wireless charging points at the edge of the field.
Bureaucracy in the field: why agri-PV is being held back by politics
Before the technology developed by TALOS can be deployed on real farms, various obstacles still need to be overcome. These
issues are tied to technology, policy and funding.
Autonomous vehicles operating in the field should ideally make no mistakes at all. Mylonas believes there is still room for improvement, particularly in terms of the software’s adaptability:
"There are plenty of scenarios that can go wrong, and I think that we are still missing the intelligence part of the robots to cope with unexpected scenarios that can occur inside a farm."
When it comes to machine approval, however, there is little cause for concern. Unlike drones, which are subject to altitude restrictions and no-fly zones, ground robots are barely regulated. A major bottleneck for the large-scale deployment of agri-PV stems from the fact that
individual EU member states have not defined what counts as agri-PV.
According to SolarPower Europe, a European trade association for the solar industry,
only five of 18 EU states have developed a legal definition.
"The policy is a bottleneck, because it really prevents the setup of bigger projects. Project developers are reluctant, since it’s really hard or time consuming to get a permit, and if you don't set up new projects, there is no testing space either." Elissen says.
In addition,
the installations require high upfront investment from farmers, who more often than not, are already struggling to stay afloat, since the solar panels must be mounted on tall structures above the fruit trees. In the Dutch pilot project, ultimately only one of the three panel configurations tested proved economically viable.
A
far more critical issue, particularly relevant in countries such as the Czech Republic, is
land ownership: there, nearly
three quarters of Czech farmland is worked by farmers who do not own it. Rent that an energy company might pay for installing solar panels, for instance, goes to the landowner rather than the farmer, leaving the latter with only EU agricultural subsidies, the
so-called CAP payments. Without a legal definition of agri-PV, it remains
unclear whether land with solar panels still counts as agricultural land, as Bím explains:
"If you go to the farmer and tell him: okay, on this field you can grow corn with the subsidies, and here I'm offering you to grow it between the modules but without the subsidy, I think they will not be interested."
At the same time, pressure is mounting: Bím reports that Czech farmers can barely find workers for fruit harvesting. An autonomous robot could therefore offer real benefits – provided there are fields where it is actually allowed to operate.
The
experts are therefore calling for clear regulatory frameworks that protect farmers and allow flexibility in choice of crops:
“There are quite some pilots that are called agri-PV but really are not, because there are too many panels and useful agriculture is not possible. Develop a sound agri-PV policy, with the farmer at the centre of the technology," Elissen demands, while Bím adds:
"We need basic rules for each member state, and there must be no limitations on crop yield. Open it for all types of crops, because every farmer grows something different." Only when it is established that subsidies and solar energy can coexist with subsidies will technical solutions such as the Husky be able to display their full potential.
Photo credit: Nicolas Congouleris / CERTH
Contacts
Project Coordinator:
Communication & press contact:
Journalist:
Website: https://talosproject.eu/
LinkedIn: https://www.linkedin.com/company/talos-eu-project/