Research Engineer, Robot Intelligence
Core
Design, develop, and integrate algorithms for robot perception, control, learning, navigation, and manipulation to create intelligent robotic systems capable of manipulation and navigation in semi-structured and unstructured environments.
Role type
Research Engineer, Robot Intelligence
Builds
Proof-of-concept/prototype robotic systems and novel user experiences leveraging Samsung's product ecosystem
Domain
Robotics, Embodied AI, Computer Vision, Reinforcement Learning
Deliverable
production ML models | product features
Required skills
Task-and-motion planning (TAMP), vision-language-action (VLA) models, open-world 3D perception, trajectory prediction, model predictive control, large-scale reinforcement learning, multimodal information fusion, whole-body control, computer vision (object detection, segmentation, tracking), multi-view image processing, kinematics, dynamics, control systems, SLAM, policy architectures
Preferred skills
System identification, real-to-sim, scene generation, sim-to-real techniques, data collection and processing for in-the-wild data aggregation, fleet management and monitoring
Technologies
PyTorch, TensorFlow, C++, ROS, CUDA
Responsibilities
Design and evaluate robotic hardware, sensors, and software suitability; Integrate sensors and actuators into prototype systems; Synthesize virtual models of robots and environments using simulation techniques; Develop and deploy data collection and model management mechanisms; Conduct pilot studies and formal experiments to benchmark performance; Maintain robot fleets and associated infrastructure
Seniority
Senior, hands-on IC
