Application Open:
Full-Time
Mohamed bin Zayed University of Artificial Intelligence (MBZUAI) is seeking a highly skilled Robotics Data Engineer with strong technical hands-on experience to join the Robot Learning Laboratory, a state-of-the-art core facility. This role is ideal for candidates who have a strong background in storage and computer engineering, with a passion for building specialized data infrastructure for robotics research. With an emphasis on practical implementation complemented by research-driven innovation, this position focuses on designing and implementing systems for collecting, processing, and managing multi-modal sensor data from robots and simulation environments.
Key Responsibilities
Robotics Data Collection and Ingestion
- Design and implement data collection systems for physical robots (mobile robots, manipulators, drones)
- Build ROS/ROS2 data recording pipelines with multi-sensor synchronization
- Develop rosbag processing tools for extraction, conversion, and validation
- Implement real-time data streaming from robots to data lake (edge-to-cloud pipelines)
- Build sensor calibration and synchronization frameworks (camera-LiDAR, IMU fusion)
- Design data collection protocols and quality standards for robotics experiments
- Develop on-robot data preprocessing and compression for bandwidth optimization
Sensor Data Processing
- Build processing pipelines for diverse sensor modalities:
- Vision: Camera images, stereo pairs, RGB-D, event cameras
- LiDAR: 3D point clouds, 2D laser scans, multi-LiDAR fusion
- IMU: Accelerometer, gyroscope, magnetometer data
- Odometry: Wheel encoders, visual odometry, GPS
- Force/Torque: Contact sensors, tactile sensors
- Implement sensor data transformations (coordinate frames, time synchronization)
- Develop data quality checks (sensor health, data completeness, outlier detection)
- Build data compression and format conversion tools (ROS → Parquet/HDF5)
- Implement sensor data visualization tools (point clouds, trajectories, video playback)
Robotics Data Infrastructure
- Design storage schemas for robotics datasets (trajectories, episodes, experiments)
- Build metadata management for robotics data (robot configuration, environment, task)
- Implement data versioning and provenance tracking for reproducibility
- Develop data indexing and search systems for efficient retrieval
- Build data annotation tools for robotics tasks (object labeling, trajectory segmentation)
- Design data sharing and collaboration workflows for research teams
- Implement data access APIs optimized for robotics workloads
SLAM and Mapping Data Management
- Process and store SLAM outputs (maps, trajectories, loop closures)
- Build map database systems (occupancy grids, point cloud maps, semantic maps)
- Implement map versioning and comparison tools
- Develop map visualization and analysis tools
- Build map merging and alignment pipelines for multi-robot systems
- Design efficient storage formats for large-scale maps (compression, tiling)
Integration with Robotics Stack
- Integrate data pipelines with ROS/ROS2 ecosystems
- Build ROS nodes for data streaming and processing
- Develop custom ROS message types and services for data platform
- Implement ROS bag replay and simulation integration
- Build bridges between ROS and data lake (Kafka, S3)
- Create ROS-based monitoring and debugging tools
- Develop ROS packages for data collection and quality checks
Real-Time Data Processing
- Build low-latency data processing pipelines for online learning
- Implement streaming analytics for robot telemetry and health monitoring
- Develop real-time anomaly detection for sensor failures
- Build event-driven architectures for robot state changes
- Implement data buffering and replay systems for debugging
- Design edge computing solutions for on-robot processing
- Optimize data pipelines for minimal latency (<100ms)
Field Support and Troubleshooting
- Provide on-site support for robotics experiments and data collection
- Troubleshoot data collection issues in real-time during experiments
- Conduct data quality validation immediately after experiments
- Train researchers on data collection best practices
- Maintain and upgrade data collection hardware and software
- Participate in on-call rotation for critical robotics experiments
- Document issues and solutions in knowledge base
Academic Qualifications Required
- Master degree in Robotics, Electrical Engineering, Computer Science, or a related field.
- A PhD degree will be preferred.
Professional Experience Required
Essential:
- 3+ years in robotics software development or robotics data engineering.
- 2+ years hands-on experience with ROS/ROS2 in production or research environments.
- Proven track record building robotics data pipelines or sensor processing systems.
- Excel in multi-tasking, organization, communication, collaboration, conflict resolution, work ethic, and time-management to thrive under pressure with a results-driven service mindset.
Robotics Expertise:
- ROS/ROS2: Deep knowledge of ROS architecture, topics, services, actions, tf, rosbag
- Sensors: Hands-on experience with cameras, LiDAR, IMU, GPS, and sensor drivers
- Sensor Fusion: Understanding of multi-sensor calibration and synchronization
- SLAM: Familiarity with SLAM algorithms and outputs (ORB-SLAM, Cartographer, RTAB-Map)
- Coordinate Frames: Strong understanding of tf, coordinate transformations, and kinematics
- Robotics Middleware: Experience with ROS message types, custom messages, and serialization
Software Engineering:
- Programming: Expert-level Python and C++; experience with ROS client libraries (rospy, rclpy, roscpp)
- Data Processing: Experience with NumPy, Pandas, OpenCV, PCL (Point Cloud Library)
- Data Formats: Knowledge of rosbag, HDF5, Parquet, Protocol Buffers
- Version Control: Proficiency with Git, GitLab/GitHub workflows
- Linux: Strong Linux skills (Ubuntu, command line, system administration)
Data Engineering:
- Pipelines: Experience building ETL/ELT pipelines with Airflow, Spark, or similar
- Storage: Familiarity with object storage (S3, MinIO), databases (PostgreSQL, MongoDB)
- Streaming: Understanding of real-time data streaming (Kafka, RabbitMQ)
- Cloud: Basic knowledge of cloud platforms (AWS, Azure, GCP)
Preferred Skills
- Advanced Robotics: Experience with mobile robots, manipulators, or drones in research or production
- Computer Vision: Deep knowledge of vision algorithms (object detection, tracking, segmentation)
- 3D Processing: Experience with point cloud processing, 3D reconstruction, or mesh generation
- Simulation: Familiarity with Gazebo, Isaac Sim, CARLA, or other robotics simulators
- Real-Time Systems: Understanding of real-time constraints and latency optimization
- Hardware: Experience with sensor hardware setup, calibration, and troubleshooting
- Machine Learning: Familiarity with ML for robotics (imitation learning, reinforcement learning)
- Open Source: Contributions to ROS packages or robotics open-source projects
- Research: Publications in robotics conferences (ICRA, IROS, RSS) or experience in research labs
- Multi-Robot: Experience with multi-robot systems and distributed robotics
