Robotics Engineer Needed: CAN Bus Motor Control, RealSense Vision, and AI-Based Autonomous Task

We are developing the upper body of a humanoid service robot and are looking for an experienced robotics engineer to help integrate motor feedback, computer vision, and AI-based autonomous task execution. -Project Overview The robot uses CAN bus actuators for its upper-body movement. The selected engineer will be responsible for reading and interpreting CAN bus data from the actuators, including motor position, movement, and status information. The engineer will then help create a system that uses this motor data, together with an Intel RealSense camera, to train or develop AI inference models capable of translating recorded movements and visual information into autonomous task execution. The goal is for the robot to learn and repeat a specific, clearly defined task reliably and autonomously. -Project Scope This project is limited to the robot’s upper body, consisting of: * Two robotic arms * The robot head * CAN bus-controlled actuators * Intel RealSense camera * AI inference and autonomous task execution There are no robotic legs involved. The robot is mounted on a four-wheel mecanum-wheel base. Base navigation is not the primary focus of this position. The main responsibility will be the manipulation and movement of the upper body. -Key Responsibilities * Connect to and read CAN bus signals from the robot’s actuators * Extract motor position, velocity, torque, current, temperature, and status data where available * Interface with the actuator manufacturer’s SDK or develop a direct CAN bus communication interface * Record and synchronize motor movement data with Intel RealSense camera data * Create a method for capturing demonstrations of the required task * Develop an AI inference or imitation-learning system that can reproduce the demonstrated task autonomously * Integrate perception, arm movement, and task sequencing * Implement safety limits, error handling, collision prevention, and emergency-stop behavior * Test and improve the reliability and repeatability of the robotic task * Document the software architecture, communication protocol, and deployment process -Required Experience * Strong experience with CAN bus communication * Experience working with robotic actuators, servo motors, or CAN-controlled joints * Experience with humanoid robots, robotic arms, or manipulation systems * Experience with Intel RealSense cameras * Strong knowledge of Python and C++ * Experience with ROS 2 * Experience with computer vision, AI inference, imitation learning, or robot learning * Experience recording and synchronizing camera and motor telemetry data * Familiarity with Linux or Ubuntu * Ability to work with existing motor SDKs and low-level CAN protocols -Preferred Experience * Experience with MyActuator motors or similar CAN bus robotic actuators * Experience using SocketCAN, USB-to-CAN adapters, CANopen, or proprietary CAN protocols * Experience with NVIDIA Jetson or similar edge-AI hardware * Experience with Orange Pi or other single-board computers * Experience with RealSense SDK, OpenCV, PyTorch, TensorFlow, or NVIDIA TensorRT * Experience creating state machines or behavior trees for robotic tasks * Experience with learning from demonstration or vision-guided manipulation * Experience developing reliable robotic demonstrations for live presentations -Expected Deliverables * Working CAN bus interface capable of reading upper-body actuator data * Real-time motor telemetry display or logging system * Synchronized recording of actuator data and RealSense camera data * AI inference or autonomous control pipeline for one defined upper-body task * Integration of the two arms, head, and camera perception * Safety and motion-limit implementation * Testing and performance documentation * Source code and setup instructions -Important Clarification This role does not involve developing humanoid walking or leg control. The robot has no active robotic legs. It uses a four-wheel mecanum base for mobility. The engineer will focus only on the upper body, specifically the two arms, head, CAN bus actuators, RealSense camera, and autonomous task execution. -Application Questions Please include the following in your proposal: 1. Describe your experience reading CAN bus data from robotic actuators. 2. Have you worked with MyActuator motors or similar CAN-controlled motors? 3. Describe a robotic arm or humanoid project you previously completed. 4. Have you used Intel RealSense cameras for robotic manipulation? 5. What approach would you recommend for translating recorded operator-controlled movements into autonomous robot behavior? 6. Do you have experience with imitation learning, reinforcement learning, behavior trees, or task-based state machines? 7. What hardware and software architecture would you recommend for this project? 8. Please provide videos, GitHub repositories, or examples of relevant previous work. We are looking for someone who can handle both low-level robotic communication and higher-level perception and autonomous task development. Please do not apply if your experience is limited only to general AI software without physical robotics, CAN bus, or motor-control experience.