Robotics Engineer - Integration & Control (m/f/d)

Prefiro GmbH; Karlsruhe; On-site; EUR 70,000 - 95,000; Full time

Benefits offered by this job

Real fields from first season

Fast path from code to robot

Stock options

Flat hierarchies

Lunch and coffee together

Job summary

Prefiro GmbH in Karlsruhe is seeking a hands-on robotics software engineer to join a small, motivated team. You will integrate robotic subsystems, implement real‑time trajectory planning, and optimize field performance of autonomous asparagus harvesters on real machines.

Requirements include a Master’s degree in a relevant field, 2+ years of robotics experience, and strong C++17 and Python 3 skills, plus ROS 2 expertise. German basics are a plus.

Qualifications

• Master's Degree in robotics, mechatronics, or related technical field.
• At least 2 years of professional robotics experience with real robots.
• Strong C++ (C++17) and Python 3 skills.
• Hands-on ROS 2 experience (Jazzy preferred).
• Experience with trajectory controllers/planning and real-time Linux.
• Solid grounding in URDF and robotics kinematics/dynamics.
• Experience with model predictive control or optimal control.
• Fluent English in a technical environment.

Responsibilities

• Integrate robots and IO into the robotics stack from startup to recovery.
• Work on MPC-based trajectory planning and real-time implementation.
• Optimize cycle times in real-world applications.
• Scale the stack to multiple robots per machine: networking and calibration.
• Tune robot computers for real-time control: Docker, CPU pinning, ROS 2 scheduling.
• Build remote access, data collection and robotics CI/CD with real field data.
• Review code and contribute to architecture decisions as a team.
• Run robots during pre-season tests and harvest season.

Skills

C++17

Python 3

ROS 2

Trajectory planning

Kinematics & dynamics

Model predictive control

CMake/colcon

Linux scheduling/latency

Git

English fluent

Education

Master's Degree in robotics or related field

Tools

MoveIt 2/Ruckig

Docker

GitLab CI

RViz/Foxglove

Job description

About Prefiro

Asparagus cost up to 25 euros per kilo this season! That's because, like most of the produce on the shelf, it's harvested exclusively by hand. Here at Prefiro, we're the first to automate these processes! We use artificial intelligence, and robotics to bring next-generation agricultural equipment to the field.

Are you excited about applying new technologies in real-world conditions, solving interesting and varied problems, and seeing firsthand the impact your ideas have?

Then join our growing team and help us deliver the first autonomous asparagus harvester to farms.

Your responsibilities

• Integrate robots and IO into the robotics stack – from startup and service to error handling and recovery.

• Work on MPC-based trajectory planning and optimal control: from approach to implementation on the robot in real time.

• Help optimize cycle times in real-world applications

• Scale the stack from one robot module to several per machine: networking, robot setup and calibration.

• Tune the robot computers for real-time control: Docker, CPU pinning, ROS 2 scheduling and multithreading.

• Build remote access, data collection and robotics CI/CD (e.g. rosbag-playback tests) to improve the robots with real field data.

• Review code, contribute to architecture decisions and debug the stack as a team.

• Run the robots during pre-season field tests and the harvest season.

What you bring

• Master's Degree in robotics, mechatronics, mechanical engineering, computer science, or a related technical field.

• At least 2 years of professional experience in robotics, including hands‑on work with real robots (not only simulation).

• Strong modern C++ (C++17) and Python 3.

• Hands‑on experience with ROS 2 (ideally Jazzy): nodes, lifecycle nodes, executors, QoS, launch files.

• Experience with trajectory controllers and trajectory planning.

• Solid grounding in robot kinematics and dynamics (URDF, Pinocchio or similar).

• Proven experience with optimal control and model predictive control for trajectory planning; reinforcement learning is a plus.

• Comfortable with CMake/colcon builds and Git on Linux.

• Experience with real‑time robot control and an understanding of soft real‑time constraints on Linux (scheduling, latency, threading).

• Fluent English, our working language.

Nice to have

• Industrial arms, especially Denso SCARA (b-CAP, RC8) or Universal Robots.

• Optimal‑control frameworks such as acados or CasADi; control barrier functions.

• MoveIt 2 or Ruckig for motion planning.

• Docker, GitLab CI and automated testing (GoogleTest, launch_testing).

• DDS or Zenoh tuning and multi‑robot setups.

• Debugging tools such as RViz, Foxglove, PlotJuggler or Nsight Systems.

• Field deployment with systemd, tmux and shell scripting.

• Basic German.

What we offer

• Real fields from your first season. Your code runs on our machines on farms in Germany.

• Short way from code to robot. New ideas usually run on the real machine within a few days.

• A problem worth solving. Green asparagus is still harvested entirely by hand.

• A small, motivated team that gets a lot done together. When you join, we're 8 full‑time, plus 2 interns and a working student.

• Open culture, flat hierarchies. We are still early, so your opinion counts in technical decisions.

• Lunch and coffee together in the kitchen in our hall.

• Employee stock options.