Lincolnshire as an unlikely robotics frontier
Drive south from Lincoln on a clear morning and the landscape flattens quickly into something almost industrial: enormous fields of brassicas, cereals and soft fruit stretching to a low horizon, broken only by drainage ditches and the occasional polytunnel. This is some of the most productive agricultural land in England, farmed at a scale that makes Lincolnshire one of the country's most significant food-producing counties. It is also, quietly, one of its most active test-beds for agricultural robotics.
The reason is not coincidence. Lincolnshire's combination of large arable and horticultural holdings, proximity to the University of Lincoln's 200-hectare Riseholme Campus — which houses research institutes, commercial partners and PhD programmes focused on agri-robotics — and a sharp post-Brexit labour shortage has made it unusually hospitable to machines being trialled on real farms, in real seasons, at commercial scale.
Yet alongside this technological momentum sits a more immediate reality: around 60,000 migrant workers once filled the UK's seasonal farm roles each year, and tens of thousands are still needed to do work that robots cannot yet do reliably or cheaply enough. So the practical question this article explores is a straightforward one — what does the arrival of these machines actually mean for the people already working the land?
What the machines actually do
The clearest way to understand what these machines do is to start with what they replace.
Robotic weeders carry cameras and processing units that scan the soil between crop rows, identifying weeds in real time using computer vision and deep learning. Some use mechanical tools; laser-weeding machines apply concentrated heat at micro-millimetre precision, killing the weed without touching the surrounding crop or leaving any chemical trace. Targeted sprayer systems, which Dyson Farming deploys on its Lincolnshire holdings, guide nozzles with precision cameras rather than blanket-spraying an entire field — the company reports cutting herbicide use by 70–90%. Across targeted sprayer systems more broadly, chemical runoff can fall by up to 95% compared with conventional broadcast application.
UV-C disease control works differently. Saga Robotics' Thorvald robot drives autonomously through crop rows at night, emitting controlled ultraviolet light that destroys fungal pathogens on the plant surface without chemical residue. In 2021, Thorvald ran commercial trials at two Kent strawberry farms — Clock House Farm (10.6 hectares) and Hugh Lowe Farm (2 hectares) — treating 7,300 linear kilometres of crop across the season. The result was a complete growing season free of fungicide sprays against powdery mildew: a documented first. The same machine is being tested at Riseholme.
Multispectral drones and distributed weather stations handle early detection. Lincolnshire Field Products uses drone fleets and on-farm sensors to flag disease or pest pressure before symptoms are visible to the human eye, enabling targeted treatment rather than precautionary spraying.
One less-obvious effect falls on labour planning. Robots tested at Riseholme under the Innovate UK-funded GRASPBERRY and FASTPICK projects count individual flowers, estimate berry weight before picking, and project ripeness timelines — giving farm managers earlier, more accurate forecasts of when and how many harvest workers they will need. The machine reshapes the workforce schedule, not just the chemical bill.
It is worth being clear about scope: the strongest documented results here relate to soft fruit. Evidence for equivalent robotic interventions in Lincolnshire's arable crops — wheat, brassicas — remains thinner.
The Riseholme research cluster
Saga Robotics did not have to come to Lincolnshire. The Norwegian company had already demonstrated Thorvald's capabilities in Scandinavian and Dutch growing environments before securing a €9.5 million Series A in 2020. It chose to base its UK operation — 35 staff — at the University of Lincoln's Riseholme Campus, and that decision is worth pausing on.
Riseholme is not simply a research park. The 200-hectare site hosts Lincoln Agri-Robotics, the Lincoln Institute for Agri-Food Technology (LIAT), and the world's first Centre for Doctoral Training in AgriRobotics, which currently has 50 PhD researchers working through applied problems in collaboration with commercial partners. For a company moving into an unfamiliar national market, proximity to that concentration of expertise, field infrastructure and soft-fruit trials is a practical pull, not a symbolic one.
The most recent addition to the cluster is JABAS.AI, a spinout launched in May 2026 from Ceres Agri-Tech — a partnership between the universities of Lincoln, Cambridge and East Anglia, backed by Research England and EPSRC. Its core offer is straightforward to explain: it allows robot fleets to navigate reliably without GPS, using lidar and computer vision instead. That matters because GPS signals fail inside polytunnels and under dense crop canopy — exactly the environments where autonomous machinery is most needed. Ceres has now launched five spinouts and, by its own count, created 34 high-value rural jobs.
The pipeline — from doctoral research at Riseholme to commercial trial to spinout to deployment — is still short and not yet proven at scale. But the infrastructure to run it is demonstrably present.
Adoption on working farms
Both Lincolnshire Field Products and Dyson Farming are large operations, and that matters for reading the picture accurately. Lincolnshire Field Products, based in South Lincolnshire, published a detailed account of its current practice in January 2026: GPS-guided machinery, data-driven soil mapping, multispectral drone fleets, and weather stations spread across multiple farm locations — all integrated and running commercially, not as a trial. The precision-targeting benefits noted in the previous section apply to Dyson Farming's selective herbicide system in the same way: cameras guide individual nozzles, the blanket spray is gone, and herbicide volumes fall by 70–90%.
What neither example answers is how much of this is spreading to Lincolnshire's smaller and family-run holdings. The county's agriculture is not dominated by a handful of large estates alone, yet the documented adoption record clusters almost entirely at that scale. Whether smaller operators are trialling these tools quietly, waiting on costs to fall, or largely unengaged is a question the county itself has not publicly answered.
The cost constraint points to why. A fully autonomous harvesting robot costs upwards of £100,000 per unit. For a farm where that sum represents a significant share of annual turnover, the business case does not close — not yet. The result is a split that characterises adoption across the sector: a farm may run drone surveillance and centimetre-accurate GPS guidance across every field while still relying on teams of seasonal workers for every hour of the harvest.
Seasonal workers caught between Brexit and automation
Those seasonal workers — the roughly 60,000 who used to arrive each year, as the opening of this piece noted — were already in a precarious position before the machines arrived. When Brexit and the pandemic converged in 2020, the flow broke sharply: Greater Lincolnshire reported over 500 unfilled driver roles alone, with harvests under threat across the county. The labour model that had underpinned decades of intensive horticulture rested on assumptions — about freedom of movement, about health — that turned out not to hold.
Greater Lincolnshire's food and farming representatives told Parliament the sector still needs a guaranteed labour supply for at least another four to five years while automation scales. That is an industry being frank about its own dependency: the technology is not arriving fast enough to replace what it is theoretically designed to replace. The Seasonal Worker Visa scheme — now drawing recruits from Bulgaria, Romania, the Philippines and India — has become the stopgap mechanism, not a long-term settlement.
The ONS has identified South Holland and the wider Lincolnshire fenlands as among the UK areas most exposed to technological job displacement, given their concentration of manual roles in planting, weeding, picking and sorting. This is a specific geographic finding, not a generalisation about Lincolnshire as a whole: it names a district where the arithmetic is particularly tight.
What that arithmetic does not easily capture is who, precisely, bears the cost. Seasonal agricultural workers in Lincolnshire are often employed through agencies on short-term contracts, housed in tied accommodation, and rarely appear in planning documents or parliamentary submissions. The voices recorded in accounts of this transition belong almost entirely to operators, researchers and industry bodies — and that is not simply a gap in reporting. It reflects how peripheral these workers are within the systems that depend on them. The distance between what automation displaces and what it creates is not closing at the same pace for everyone; those least positioned to cross it are also the least visible in the conversation about how to manage it.
The skills gap the machines are opening
The jobs automation is generating are not the same jobs being displaced. AI systems management, robot maintenance, software engineering and data analysis all require qualifications and training routes that most seasonal agricultural workers do not hold and cannot easily acquire. The five Ceres Agri-Tech spinouts — which, as noted earlier in this piece, have together created 34 high-value rural jobs — represent a genuine contribution to the county's rural economy, but not a transition pathway. Against a seasonal workforce counted in tens of thousands across Lincolnshire's fields, the arithmetic is blunt.
Riseholme is training the next generation of agri-robotics specialists, but at PhD level. That is exactly the right investment for advancing the technology; it is not a vocational route accessible to a South Holland agency worker whose weeding contract ends in August. The pathway from seasonal harvest labour to robot technician does not currently exist as a structured offer anywhere in the county that the public record shows — and that is not an oversight. It is a structural feature of how the transition is being managed: investment flowing into research and commercialisation, workforce retraining left largely unaddressed.
Greater Lincolnshire's farming sector told Parliament it needs a secure labour supply for four to five more years — a window that runs to roughly 2029 or 2030. That is also the window in which the shift could be shaped rather than simply absorbed. Lincoln College runs agricultural courses; South Kesteven District Council and Lincolnshire County Council both carry rural employment within their remit. Whether any of those institutions is treating the gap between current seasonal roles and emerging technical ones as a problem they own is the question the county has not yet publicly answered.