Senior Engineer, State Estimation (R5784)
San Diego, CA, USA
Job Description:
Join Shield AI’s Hivemind SDK State Estimation and Vision team to build the tools that help engineers develop, understand, and validate autonomous navigation systems. You will turn complex sensor data and estimator behavior into reproducible experiments, clear diagnostics, and actionable performance insights.
Your work will span data ingestion, sensor replay, visualization, calibration workflows, automated evaluation, and regression testing. Working closely with state estimation, vision, simulation, and flight test engineers, you will shorten the path from discovering an issue in real-world data to understanding its cause and validating a fix.
What you'll do:
- Build and maintain tools for ingesting, inspecting, synchronizing, and replaying camera, IMU, GNSS, and other navigation sensor data.
- Develop visualization and diagnostics for trajectories, coordinate frames, uncertainty, estimator residuals, feature tracks, and sensor health.
- Create reproducible evaluation workflows that track datasets, software versions, calibration, and configuration.
- Build automated benchmarks and regression tests for estimation accuracy, drift, robustness, runtime, and resource usage.
- Enable batch experiments, parameter sweeps, and comparisons across algorithm versions, datasets, and operating conditions.
- Help engineers diagnose failures involving timestamp alignment, calibration, frame conventions, data quality, and estimator behavior.
- Integrate tooling into SDK APIs, command-line interfaces, and continuous integration workflows, with clear documentation and examples.
- Partner with tool users to identify workflow bottlenecks and deliver reliable, maintainable solutions.
Required qualifications:
- A degree in computer science, robotics, electrical engineering, aerospace engineering, or a related field, or equivalent practical experience.
- Strong software engineering skills in Python and modern C++, with the ability to work effectively in both.
- Experience building developer tools, data processing pipelines, test infrastructure, or analysis software for robotics, autonomy, computer vision, or a related technical domain.
- Working knowledge of linear algebra, probability, 3D geometry, and coordinate transformations.
- Familiarity with state estimation concepts such as sensor fusion, filtering, nonlinear optimization, visual odometry, or SLAM.
- Ability to work with time-series and sensor data, including timestamps, units, reference frames, missing measurements, and data quality.
- Experience developing on Linux using version control, automated testing, debugging, and profiling tools.
- Ability to design usable interfaces, write maintainable code, and communicate technical findings clearly.
Preferred qualifications:
- Experience with visual-inertial odometry, SLAM, factor graphs, bundle adjustment, Kalman filtering, or navigation in GNSS-degraded environments.
- Experience with camera and IMU calibration, temporal synchronization, sensor noise characterization, or ground-truth alignment.
- Familiarity with libraries such as Eigen, OpenCV, Ceres, or GTSAM.
- Experience with robotics data formats and visualization tools such as ROS/ROS 2, MCAP, Foxglove etc.
- Experience building interactive 2D/3D visualization, dashboards, or browser-based engineering tools.
- Experience with simulation, software-in-the-loop or hardware-in-the-loop testing, and flight-log analysis.
- Familiarity with dataset versioning, experiment tracking, distributed evaluation, or large-scale data processing.
- Experience profiling numerical software on embedded or resource-constrained systems.
- Experience shipping SDKs, Python bindings, reusable libraries, or tools used by external developers.
160000 - 250000 USD a year