Agricultural robotics offers important opportunities for addressing global food security, as approximately one in eleven people face hunger worldwide, while agriculture must operate under increasing constraints on water, labor, and land.
Agricultural manipulation, from interactive plant inspection to fruit harvesting, requires robots to understand and operate under substantial variability caused by dynamic outdoor conditions, such as lighting and wind, and by plant morphology, including differences in growth stage and genotype. This variability presents a distinct challenge for robot learning. Unlike home or industrial settings, where objects, poses, and environmental conditions are often structured or known, agricultural environments involve significant variation even among plants of the same species, including differences in geometry, pose, size, maturity, and spatial arrangement.
Can data-driven methods scale to agricultural robotics despite orders-of-magnitude greater variation than warehouse and household domains? This workshop seeks to explore merging robot learning research for agricultural robotics by bringing together researchers, students, industry stakeholders, and practitioners to re-examine these challenges and opportunities.
Our workshop features speakers from academia and industry, with a variety of backgrounds relevant to our themes. To the best of our knowledge, this will be the first such workshop at CoRL. However, we believe that we are at an inflection point where the field of robot learning continues to advance and veritably solve tasks across fields, and the challenges in real-world agriculture have become more pertinent. This workshop will bring stakeholders together to discuss and consider the possibilities at this inflection point.
We solicit submissions and discussion on questions including, but not limited to:
Chief Technology Officer, Blue River Technology
Chief Technology Officer of Blue River Technology (acquired by John Deere), which applies robotics to precision agriculture for real-time, AI-driven weed removal and crop management. His deep industry experience will help motivate new research problems and bridge the gap between the research and industrial communities.
UC Berkeley — AUTOLab
For three decades, his lab has carried out research in automation and manipulation across a range of robot applications, with agricultural robotics as an important component — including precision irrigation, pruning and irrigation in precision polyculture farming, and plant growth modeling. He will share insights into the challenges of bringing data and learning to agricultural robotics.
UC Davis — Plant Simulation Laboratory
His lab developed Helios, a leading simulation and modeling platform for plants that supports 3D modeling and GPU acceleration. He will discuss how simulation enables data curation and robot learning for agricultural robotics.
Katie Driggs-CampbellJoining online
University of Illinois Urbana-Champaign — Human-Centered Autonomy Lab
Affiliated with the Center for Digital Agriculture, her lab has worked on improving autonomy in agricultural robotics, including managing multiple robots at scale and failure recovery. She will discuss the uncertainties of unstructured outdoor environments and the opportunities offered by methods ranging from model-based approaches to data-driven techniques.
Half-day workshop, 8:30 AM – 12:30 PM (local time, TBD). The workshop will be held in person, with supplemental online participation supported for up to 50 remote attendees.
| Time | Session |
|---|---|
| 8:30 – 8:40 AM | Welcome — workshop objectives and format |
| 8:40 – 9:20 AM | Invited Talks (2 × 15-minute presentations + 5-minute Q&A each) |
| 9:20 – 9:50 AM | Accepted Talks / Poster Blitzes |
| 9:50 – 10:30 AM | Invited Talks (2 × 15-minute presentations + 5-minute Q&A each) |
| 10:30 – 11:00 AM | Coffee Break (Poster Sessions & Demos) |
| 11:00 – 11:30 AM | Panel Discussion |
| 11:30 AM – 12:15 PM | Breakout Discussions / Brainstorming Sessions |
| 12:15 – 12:30 PM | Concluding Remarks, Action Items, Closing |
Schedule is tentative and subject to change closer to the workshop date.
We invite contributions from agriculture researchers applying learning-based methods aligned with any of the topics above. Submissions should be up to 3 pages (excluding references) and will be treated as non-archival.
Papers will be selected based on their relevance to the workshop themes and their potential to seed meaningful discussion among attendees, with reviews informing the final selection (up to 10 papers total). In line with CoRL 2026 policy, we will not accept papers that have already been published at, or accepted to, the main CoRL 2026 conference. Speakers and participants may, however, highlight relevant agriculture robotics papers appearing at CoRL 2026.
Accepted papers will be spotlighted at the workshop, followed by a poster session, and will be made available on this website. Authors are encouraged to participate in the breakout discussions.
In line with CoRL 2026 requirements, we will release a post-workshop artifact — a position paper — approximately one month after the workshop. The position paper will be authored by the organizers, drawing on the invited talks, spotlight presentations, and breakout discussions, and incorporating insights from speakers, panelists, and participants. We intend to submit the resulting position paper to a venue such as Science Robotics.