Atlas Staff Perception Engineer, Tactile Sensing

Job description

Job responsibilities

• Collaborate with multidisciplinary teams to design, evaluate, and demonstrate tactile sensing data representations that enable reliable closed-loop robotic manipulation and interaction with the environment
• Architect and implement highly optimized data pipelines to acquire, process, and stream high-bandwidth data from tactile sensors
• Design and deploy automated data collection routines using robot end-effectors or instrumented test rigs to generate ground-truth datasets for sensor calibration and modeling
• Design and deploy automated data collection routines using robot end-effectors or instrumented test rigs to generate ground-truth datasets for sensor calibration and modeling
• Conduct hands-on testing to characterize sensor behavior under operational loads including motion, vibration, and thermal variation
• Develop data processing pipelines that leverage advanced DSP techniques to filter, fuse, and compress raw sensor signals into optimized representations used by robotic perception and behavior models
• Model sensor performance and drive the integration of tactile sensing into simulated virtual environments (e.g., Isaac Sim, MuJoCo) to effectively bridge the sim-to-real gap
• Work closely with software, systems, manufacturing, and test engineering teams to define hardware interfaces and ensure successful system integration

Job requirements

• MS or PhD in Electrical Engineering, Computer Science, Robotics, or related field. BS with significant relevant experience will also be considered
• 5+ years of experience developing embedded systems in robotics, automation, or similar electromechanical products
• Strong understanding of the physical principles behind tactile sensing and the ability to model sensor behavior and translate physical limitations into robust data representations
• Experience using version-control (Git) and developing software in Python and C++ for robotic systems, emphasizing high-bandwidth data acquisition.
• Proficiency in Python and deep learning frameworks such as PyTorch applied to sensor data processing or representation learning
• Strong technical communication and documentation skills
• Experience designing and executing hardware-based data collection campaigns to generate ground-truth datasets for sensor calibration, modeling, or simulation environments such as Mujoco or Isaac Sim
• Experience integrating multiple sensor modalities such as vision, force/torque sensors, IMUs, or tactile arrays into complex electromechanical systems to build cohesive environmental representations.