Computer Vision R&D Engineer / Engineerka

SORIGO Sp. z o.o. is an IT company with over a decade of experience serving large clients. The company focuses on innovative IT solutions, employee development, and R&D projects, particularly in UAV navigation in GNSS-denied environments.

• Experience in Computer Vision, Robotics, or State Estimation, especially in R&D projects.
• Very good knowledge of C/C++ and Python.
• Experience with Linux, Ubuntu, and ROS.
• Knowledge of Visual Odometry (VO) / Visual-Inertial Odometry (VIO).
• Familiarity with feature tracking, loop closure, and visual place recognition.
• Practical knowledge of Kalman Filter, EKF, and sensor fusion using camera, IMU, barometer, and magnetometer.
• Experience with ArduPilot EKF3 and external navigation data interfaces.
• Experience profiling and optimizing algorithms on ARM platforms and real-time environments.
• Experience working with Raspberry Pi or similar edge devices.
• Experience analyzing flight data such as ROS bags.
• Ability to design experiments, prepare test scenarios, and statistically analyze results.
• Experience with camera–IMU synchronization, sensor calibration, vibration handling, and EKF tuning.
• Documentation skills for experiments, methods, parameters, results, and conclusions.
• Scientific publications, PhD or ongoing PhD, or R&D project experience.

Nice to have:

• Experience with Deep Learning and Edge AI models for feature extraction/matching.
• Experience with SLAM, loop closure, or visual place recognition.
• Experience limiting long-term position drift.
• Knowledge of ATE / RPE metrics and benchmarking localization algorithms.
• Optimization of Computer Vision algorithms on edge/ARM devices.
• Experience with UAV, drones, or navigation without GNSS.
• Testing algorithms under challenging conditions like low light, motion blur, vibration, altitude changes, high speed, or low terrain texture.

• Develop and test Computer Vision and Visual-Inertial Navigation algorithms.
• Work with C/C++, Python, Linux/ROS, Raspberry Pi 5, and ArduPilot.
• Improve algorithm robustness against poor lighting, motion blur, vibrations, variable altitude, and low terrain texture.
• Develop methods to reduce position drift during long UAV flights.
• Analyze real UAV flight data and benchmark localization algorithms.
• Design and conduct experiments and tests in demanding environments.
• Collaborate with R&D team and participate in test campaigns on test ranges.
• Work on solutions using Edge AI and Deep Learning.
• Contribute to the development of GPS/GNSS-denied navigation systems.

Applicants must include a consent clause for personal data processing in their CV as per Polish data protection law.