Job Description

Develop, test, and deploy autonomy algorithms for humanoid robots—covering localization & mapping (SLAM/State Estimation), motion planning, whole-body control, and operational safety in complex, dynamic environments.

Key Responsibilities:

• Develop, test, and deploy advanced localization and mapping algorithms tailored for humanoid robots, calibrating and integrating data from LiDARs, cameras, IMUs, encoders, foot sensors and other sensors.

• Develop, test, and deploy advanced motion planning and control algorithms tailored for humanoid robots to enable safe, smooth and time-optimal navigation in complex, dynamic environments.

• Collaborate with other team members to determine the types of sensors, their placement, and field of view to ensure safety during robot navigation.

• Collaborate with cross-functional teams to design data collection and evaluation pipelines, including defining metrics and validation frameworks.

• Deliver robust, real-world-ready software that meets high reliability standards.

• Contribute to the continuous improvement of the autonomy stack in collaboration with other engineering teams.

Required Qualifications:

• BS+ in Robotics/CS/EE/ME or related.

• Strong C++ and Python; experience with ROS/ROS2, Linux, Git.

• Solid state estimation: EKF/UKF, VIO/VINS, graph-based SLAM (factor graphs/BA).

• Motion planning exposure: A*/D*, PRM/RRT*, trajectory optimization, basic MPC.

• Controls foundation: kinematics/dynamics; basic to intermediate whole-body control (QP/CLF-QP).

• Multi-sensor handling: LiDAR, cameras (stereo/RGB-D), IMU; time sync & calibration.

Preferred Qualifications:

• Humanoid experience (footstep/contact planning, centroidal dynamics, impedance).

• Hands-on with GTSAM, Kimera, Cartographer, ORB-SLAM/DSO; ceres, Sophus, Eigen.

• Advanced planning/control: whole-body MPC, QP cascades, time-optimal/safe corridors.

• Simulation (Gazebo/Ignition, Isaac Sim, MuJoCo), log replay/rosbags.

• Performance tuning (profiling, lock-free, zero-copy, SIMD) and basic CUDA.

• Safety engineering exposure; AI/RL/IL integration for loco-manipulation is a plus.

Tools & Stack

• ROS2, C++, Python, Eigen, ceres-solver, GTSAM, Sophus

• OpenCV, PCL, pybind11, protobuf/ROS msgs

• Gazebo/Ignition, Isaac Sim, MuJoCo

• Git, CMake, clang-tidy/format, GoogleTest, Docker/CI