Robot Simulation and 3D Scene Engineer

Job Responsibilities

• Responsible for building robot simulation environments and digital twin scenarios, supporting robots in perception, navigation, operation, and task verification in complex indoor and outdoor environments.
• Responsible for 3D modeling, scene editing, physical property configuration, and simulation asset management for typical scenarios such as roads, parks, public facilities, and underground garages.
• Responsible for building a reusable simulation scene library, including common environments and operational objects such as ground, walls, columns, cabinets, doors, boxes, obstacles, trash bins, cones, barricades, vehicles, etc.
• Responsible for sensor configuration and data generation in the simulation environment, including basic modeling and data export for sensors such as RGB cameras, depth cameras, LiDAR, IMU, and thermal imaging.
• Responsible for configuring and randomizing physical parameters in simulation scenarios, including lighting, materials, occlusion, target positions, object poses, friction, collisions, noise, and other factors, to enhance the algorithm's generalization ability to real-world scenarios.
• Responsible for aligning the robot simulation model with the real robot parameters, including dimensions, joint limits, end-effector tools, collision bodies, and sensor installation poses.
• Validate robot perception, SLAM, motion-control, and task-planning modules in simulation, including reachability analysis, collision detection, path validation, and edge-case testing before deployment to physical robots.
• Participate in some 3D vision-related work, including point cloud processing, depth map processing, 3D object annotation, pose estimation assistance, camera calibration, and 3D data quality inspection.
• Responsible for the alignment analysis between simulated data and real collected data, and assisting in the construction of datasets for target recognition, scene understanding, and operation verification.
• Responsible for the simulation toolchain, asset management standards, data export processes, and test case accumulation, improving the team's simulation verification and data generation efficiency.

1. Basic General Ability (Base)

• Familiar with Linux/Ubuntu development environment, with good engineering development and debugging skills.
• Proficient in Python, with certain C development capabilities, able to complete the development of simulation scripts, toolchains, and data processing workflows.
• Familiar with one or more tools in the robotic simulation, 3D modeling, or game engine toolchain.
• Familiar with common 3D data representations, including mesh, point cloud, depth map, pose, camera intrinsic/extrinsic parameters, coordinate system transformation, etc.
• Understand the basic coordinate systems of robots, familiar with concepts such as TF, pose transformation, camera extrinsics, and 3D spatial relationships.
• Possess strong scene reconstruction ability, capable of building reasonably structured, operable, and reusable simulation environments according to real test requirements.
• Have good engineering practices, able to maintain simulation assets, scripts, configuration files, and documentation, ensuring scenes are reproducible in collaborative environments.
• Possess a basic understanding of robotic systems and can collaborate with modules such as perception, SLAM, motion control, and systems engineering.

II. Specialized Technical Directions

Track A: Robot Simulation and Digital Twins

• Familiar with one or more simulation platforms such as Isaac Sim, Gazebo, MuJoCo, PyBullet, Unity, Unreal Engine, Webots, CoppeliaSim.
• Able to import, configure, and debug robot models, sensor models, environment models, and task object models.
• Familiar with one or more common model formats such as URDF, MJCF, USD, SDF, FBX, OBJ, GLTF.
• Able to configure simulation properties such as rigid bodies, colliders, joint constraints, friction, mass, inertia, and sensor parameters.
• Able to build testable simulation scenarios according to robot task requirements, and not just create static 3D displays.
• Experience in robot navigation, robotic arm operation, mobile operation, or humanoid robot simulation is preferred.

Track B: Complex Scene Modeling and Asset Management

• Has experience in building 3D scenes, able to complete modeling or asset organization of scene elements such as roads, interiors, garages, campuses, cabinets, doors, and obstacles.
• Familiar with one or more modeling tools such as Blender, Maya, 3ds Max, SketchUp, or Houdini.
• It can restore a simulation environment based on real-world photos, CAD, point clouds, or simple measurement data.
• Able to create or organize a reusable scene asset library, and standardize naming, hierarchy, materials, colliders, and scale units.
• Familiar with the effects of factors such as lighting, material, reflection, shadows, and occlusion on visual algorithms and sensor simulation.
• Experience in procedural scene generation, Domain Randomization, and synthetic data generation is preferred.

Track C: 3D Vision and Sensor Data Processing

• Familiar with one or more 3D vision or data processing tools such as OpenCV, Open3D, PCL, NumPy, PyTorch3D.
• Familiar with camera models, depth maps, point clouds, extrinsic calibration, coordinate transformation, and 3D projection relationships.
• Able to handle RGB-D data, point cloud data, LiDAR data, or multi-camera data.
• Able to support tasks such as target 3D position estimation, object pose estimation, scene geometry analysis, and operable area determination.
• Capable of assisting perception algorithm engineers in simulation data generation, data labeling, data quality inspection, and Sim-to-Real discrepancy analysis.
• Experience in robotic vision, autonomous driving perception, AR/VR, 3D reconstruction, NeRF / Gaussian Splatting is preferred.

Preferred Qualifications

• Experience in building a robot simulation platform from scratch.
• Has practical project experience with Isaac Sim / Omniverse / Gazebo / MuJoCo / Unity / Unreal.
• Has experience with mobile robots, robotic arms, humanoid robots, quadruped robots, or autonomous driving simulations.
• Has experience in robot operation task simulation, such as grasping, pushing doors, pressing buttons, handling, tool operation, obstacle clearing, etc.
• Has experience in modeling complex indoor and outdoor scenes, such as underground garages, campus roads, buildings, factories, warehouses, and urban blocks.
• Experience in synthetic data generation, simulation data annotation, domain randomization, and Sim2Real.
• Familiar with ROS / ROS2, capable of integrating simulation environments into robotic algorithm pipelines.
• Familiar with the simulation and calibration processes of sensors such as cameras, depth cameras, LiDAR, and IMU.
• Has experience with 3DGS, NeRF, 3D reconstruction, point cloud registration, and SLAM map reconstruction.
• Has experience with high-quality open-source projects, robotics competitions, autonomous driving simulations, and digital twin projects.
• Experience in AI-First development, familiar with tools like Claude Code / Codex to improve R&D efficiency.