The KAFE Challenge: When Robotics Makes Agroecology Profitable
- May 22
- 5 min read
Where Does the KAFE Challenge Come From?
The KAFE Challenge — Kubota Automation For Ecology — was born from a conviction: agricultural robotics can only reach its full potential if it fundamentally transforms crop management practices.
Without this transformation, it remains a simple technical optimization. With it, it becomes a true driver of agroecological transition. This conviction steered the project toward an ambitious goal: exploring complex agronomic practices in order to demonstrate their added value.
The central question is simple to formulate: how can a complex agroecological approach be made economically and technically viable through robotization?
Robotics must be able to transform our crop management and farming methods, to do better in less time, and thus reveal its full value." Christophe Aubé — CEO, Agreenculture
The answer proposed by the project is based on fully robotized strip cropping, carried out with a standard tractor and standard implements (seed drill, sprayer, fertilizer spreader, etc.), operated autonomously.
The goal is not to revolutionize the tools, but to revolutionize the way they are used.
Strip Cropping: Diversity, Boundaries, and Precision
Strip cropping consists of growing different species in parallel strips on the same plot. Each strip is sown, cultivated, and harvested at different times, according to precise agronomic rules.
This technique already exists without robotics, notably in the Netherlands — but it remains very marginal there. Why? Because it multiplies the interfaces between crops, and each interface means time, vigilance, and precision. In short: complexity. And in the agricultural world, complexity translates directly into workload for the farmer.
Strip cropping maximizes the number of boundaries between crops. Each boundary imposes high requirements — particularly in terms of spraying, where even slight product drift can impact an adjacent crop. This calls for a level of precision agriculture that only robotics can achieve systematically and reliably.
The project's equation is therefore as follows: we accept increased complexity in crop organization, precisely because that complexity has no impact on the farmer's schedule. It is the robotics that absorbs this time — and therein lies the fundamental value of autonomy.
Technical Challenges to Solve
Moving from vision to reality required solving several concrete problems, closely tied to field conditions and farm machinery.
Compatibility of implements with spatial organization
The definition of strip cropping bands depends directly on the working widths of the implements used: seed drill, sprayer, fertilizer spreader. The width of the strips must be an exact multiple of the implement widths, to avoid any "underperformance" (using only a third of a seed drill's width, for example, would be economically absurd). This requirement became a major challenge when the available implements were only confirmed at the very last moment of the KAFE Challenge.
Irrigation and economic consistency
The plots involved were irrigated. Irrigating over 20 meters has a different economic logic than irrigating over 80 meters. The project had to integrate this constraint into the strip definition to maintain the agronomic and economic coherence of the operation.
Precision agriculture and density adjustment.
Soil analyses had produced seeding density recommendations by zone. The challenge was to adjust these densities in real time, while accounting for the fact that different implements (seed drill, row cultivator, mechanical tool) had to pass over the same location in sequence.
Faced with uncertainty about the implements, the solution adopted was to use the geographic coordinates of the crop rows as a fixed reference. Although this introduced a slight sub-optimization, it preserved spatial consistency regardless of the implements available — an approach similar to that used in the 2018 Centeol Challenge.
A Five-Partner Project: Governance as a Key Issue
The KAFE Challenge is not a unilateral technology project. It relies on an ecosystem of five partners whose alignment is a prerequisite for success.
Partner | Role |
The farmer | Soil knowledge, landowner |
The seed supplier: RAGT | Crop selection, seed choice, and sowing schedules |
Agreenculture | Machine and implement autonomy |
Kubota | Tractor manufacturer |
Kverneland / Kubota | Machinery supplier, implement provider |
The agroecological approach only makes sense if all these stakeholders are aligned on the final outcome, and committed over multiple years. Implementing strip cropping for a single season offers limited benefit. The full benefits — long rotations, biodiversity preservation, nitrogen-trap crops, resilience to climate variability and pests — only materialize over several years.
The project also raised the question of the relationship between partners: no longer a client/supplier logic, but a genuine project partnership. Informal meetings held throughout the season helped bring machinery and implement manufacturers along in understanding what autonomy truly means, and to collectively identify the development work needed to achieve the ecological ambition through robotics.
Our ambition is to build a true synergy with our partners, well beyond a conventional client/supplier relationship." Christophe Aubé — CEO, Agreenculture
Successes, Frustrations, and Lessons Learned
The project delivered rich lessons, both in what it achieved and in what it was unable to demonstrate.
What worked
✓ Partners fully grasped the value of strip cropping while also appreciating its complexity, making it possible to precisely identify the additional development work required.
✓ A wide range of different implements were robotized and successfully deployed, demonstrating the compatibility of conventional farm machinery with an autonomous approach.
✓ Real-world use of the system — even at an accelerated level of complexity — produced a specification document grounded in field facts rather than assumptions.
✓ Partners gained a genuine understanding of what autonomy truly means, far beyond passive auto-steering.
Challenges and frustrations
! A project limited to a single season. Agroecology, and strip cropping in particular, must be measured over multiple years. Without a multi-year commitment, it is impossible to demonstrate benefits on biodiversity, long rotations, or climate resilience.
! Significant wildlife damage. Wild boar caused considerable disruption to the plots, making thorough and reliable measurement of agronomic performance impossible.
! Late confirmation of implements. Not knowing the available implements in advance introduced constraints in strip definition. This issue is fixable, but highlights the importance of early-stage coordination.
! Spring crops only. Combining winter and spring crops would have amplified the positive effects on biodiversity — an identified area for improvement in future editions.
What's Next?
The ambition for the next iteration of the challenge is clearly stated: go further, broader, and longer.
Objectives for the next edition
Multiple machine types: interoperability of robotic solutions will be a key factor
Larger and more diverse plots to test at greater scale
A multi-year commitment, a necessary condition for measuring biodiversity and resilience
Demonstrate profitability from the first harvest year for the farmer
Measure biodiversity using a rigorous scientific protocol: several times per year, over multiple years, at various points across the plots
The project is built on a strong conviction: biodiversity cannot be measured on paper once a year. It requires a rigorous technical and economic protocol, regular surveys (monthly, at different locations), and an observation period spanning multiple seasons. Only under these conditions will robotized strip cropping be able to demonstrate a dual return on investment: economic and environmental.
"Our objective is a dual ROI: economic and environmental. Demonstrating profitability from the first harvest year, and over multiple years, the preservation or restoration of biodiversity." Christophe Aubé — CEO of Agreenculture
The KAFE Challenge has proven that such a vision is technically achievable with standard farm machinery. What remains is to demonstrate it at scale, over time, and with multiple partners around the table.
