AI/Robotics Engineer - Learning-Driven Manipulation for Agrifood & Industrial Robotics
Role details
Job location
Tech stack
Job description
- Develop dexterous manipulation policies for complex, uncertain environments in both Agrifood and Manufacturing.
- Integrate Transformers, VLAs, Diffusion Policies, Flow Matching, Foundation Models for Robotics (RT-X style), and Imitation Learning into real robot systems.
Build full AI pipelines, from data to deployment
- Design and execute data collection campaigns in real environments (greenhouses, orchards, industrial settings).
- Lead data curation, annotation pipelines, dataset augmentation, and multimodal dataset preparation (RGB-D, proprioception, tactile, force-torque data).
- Train behaviour cloning, reinforcement learning, and foundation-model-prompted RL/IL.
- Deploy learned models on-robot / on the edge, optimizing for latency, reliability, and real-time performance under field conditions.
Advance perception for challenging environments
- Develop 3D perception, crop/fruit detection, pose estimation, grasp synthesis, and scene understanding for unstructured natural environments.
- Fuse vision, proprioception, and force data to enable safe physical interaction.
Make real robots learn and improve
- Integrate your algorithms with our robotic platforms (mobile manipulators, cobots, custom agrifood systems).
- Validate your solutions both in the lab and on field deployments.
Collaborate, learn, and grow with us
- Work with an international consortium of companies, R&D partners, and farmers/growers.
- Contribute to scientific publications, demos, project deliverables, and client interactions.
- Participate in project planning, execution, and new proposal definition.
Requirements
Are you excited about shaping the next generation of AI-powered robotic manipulation in real, unstructured environments?
The Robotics and Automation Unit at Eurecat Technology Center (Barcelona) is looking for a passionate engineer or researcher who wants to push the boundaries of robot learning, dexterous manipulation, and perception-rich autonomy in agriculture and advanced manufacturing., Must-Have
- MSc or equivalent experience in Robotics, AI, Computer Vision, Mechatronics, or related fields.
- Practical experience in robotic manipulation, perception, or robot learning (projects, internships, theses, or professional work).
- Strong programming skills in Python and experience with deep learning frameworks (e.g., PyTorch).
- Experience with ROS or ROS 2, robot control, motion planning, or sensor integration.
- Ability to work both independently and collaboratively in multidisciplinary environments.
- Motivation to build things end-to-end: experimenting, training, and deploying on real robots.
Nice-to-Have
-
Knowledge of Transformers, Diffusion Policies, Flow Matching, VLA models, or multimodal robot foundation models.
-
Experience in Behaviour Cloning, Imitation Learning, Offline RL, or Deep RL.
-
Experience training large policies or models.
-
Skills in 3D vision, segmentation, tracking, or pose estimation.
-
Experience with C++, motion planning, or force-based interactions.
-
Hands-on experience in Agrifood Robotics (harvesting, sorting, detection of fruits/crops).
-
Publications, open-source contributions, or robotics competition experience.
-
Videos, demos, or project repositories showcasing your work.
Benefits & conditions
- Hybrid work (home office/ work in the office).
- Flexible Schedule.
- Shorter workday on Friday and Summer Schedule.
- Flexible remuneration package (health insurance, transport, lunch, studies - training and kindergarten).
- Eurecat workers can join the Eurecat Academy courses.
- Eurecat workers can attend to our weekly thematic sessions to be aware of Eurecat areas of knowledge.
- Language's training (EN, Catalan and Spanish).