We are seeking an exceptional AI / Robotics research engineer to design and build vision -based independent systems that operate in the transmitted GPS environments, rapidly variable, high -speed (for example, drones, air robots). It will work on intersection of deep learning, actual time vision, typical alert control (MPC), robots, creating new ML models and systems-not only to sew open source libraries.

You will collaborate with our bachelor’s devices, systems, and Mechatronic teams to spread artificial intelligence models in the episode-in the real world settings with real consequences.

Tasks

• Design and implementation of automatic learning structures for mobility and vision -based control.
• Development and improvement of real -time perception systems using CNNS, transformers or vision based on events.
• Apply deep reinforcement learning and MPC to direct independent behaviors in dynamic environments.
• Strong architects engineer tubes fusion sensors with IMUS, events cameras, Lidar, radar, and optical medals measuring.
• Building, training and evaluating ML models in simulation and spreading them on the included devices.
• Merging artificial intelligence with the traditional control theory to achieve the independence of the closed loop on a large scale.
• Use and improve NVIDIA (Cuda, Tensorrt, Deepstream, Jetson) tools for performance.

requirements

• 3-7+ years of practical experience in AI/ML, robots, or independent systems research (industry or doctoral/postdoc).
• Deep knowledge of the real -time computer vision, sensor integration, and models -based control.
• Strong experience in deep reinforcement learning, optimal control,/or hybrid artificial intelligence systems.
• A busy record for design of allocated ML models or training pipelines, not just reusing open source structures.
• Advanced python and C ++ efficiency; Deep familiarity with pytorch and/or tensorflow.
• Practical experience with NVIDIA GPU tool groups to spread the artificial intelligence or actual time.
• Experience in publishing in the simulation and devices in the episode.
• It is preferable to have a space experience, defense or other high -speed robots.

The reward points

• Experience with events -based cameras (for example, bewildering, dvs).
• Publications in the first -class AI/Robotics Conferences (for example, ICRA, NeuPs, CVPR).
• Knowing GNC systems, actual time operating systems (RTOS) or airline software.
• Previous start -up background, defense or airline engineering.

benefits

• Be part of the team of building independent systems from A to Z.
• Access to devices and simulation environments.
• Competitive salary + property rights.
• Full -time access, at the site to our laboratory space in Berlin TXL.