About the Role
The role focuses on building and maintaining in-house real-time 3D tracking that anchors CAD models to physical steel structures in dynamic shop-floor environments. You will advance existing internal systems, taking full ownership of architecture, optimization, and deployment across AR tablets and headsets used daily in industrial settings.
Responsibilities
- Lead end-to-end development of the on-device object tracking core, spanning design, implementation, and deployment
- Develop real-time 6DoF pose estimation and continuous tracking for textureless steel components using CAD models as reference
- Integrate object tracking with visual-inertial odometry, depth sensors, and IMU data, including handling tracking loss, drift correction, and recovery
- Optimize C++ vision pipelines for iOS and Apple silicon, ensuring stable performance on iPad and headset hardware under CPU/GPU constraints
- Scale tracking to large industrial structures by managing partial visibility, multi-region strategies, and preprocessing CAD meshes from IFC/STEP exports
- Bridge native C++ modules with Unity through plugins, collaborating closely with the Unity team from prototyping to release
- Establish performance metrics for tracking quality and validate them under real-world conditions such as occlusion, variable lighting, and clutter
Requirements
- Minimum of five years of professional experience in computer vision or related fields with emphasis on real-time systems
- Successfully shipped at least one real-time computer vision feature on a mobile platform, including iOS or Android
- Proficient in modern C++; Python knowledge is a plus
- Strong understanding of 3D geometry concepts including camera models, calibration, projective geometry, and linear algebra
- Experience with nonlinear least-squares optimization methods such as Gauss–Newton or Levenberg–Marquardt and SE(3) pose parameterization
- Hands-on background in pose estimation and object tracking algorithms
- Proven experience integrating native C++ code with Unity or similar native-to-mobile plugin environments
- Skilled in performance profiling and optimization on ARM architectures, including frame budgeting and tools like Instruments
- Familiarity with OpenCV for computer vision tasks
Nice to Have
- Experience with model-based 3D object tracking using region-based, edge-based, or hybrid techniques
- Knowledge of Metal and GPU compute programming on Apple silicon platforms
- Background in visual-inertial odometry or SLAM systems, including ARKit or ARCore
- Experience in state estimation and multi-sensor fusion using Kalman filters, factor graphs (e.g., Ceres, g2o, GTSAM), extrinsic calibration, and time synchronization
- On-device machine learning deployment using Core ML, ONNX Runtime, or Unity Sentis, particularly for pose refinement
- Handling CAD data including mesh processing, decimation, and working with IFC/STEP-style file exports
- Industrial vision, robotics, or augmented reality domain experience
About the Role
The software overlays AR work instructions onto structural steel during fabrication, running on tablets and headsets across North American shop floors. It guides workers through fit-up, welding, and inspection using CAD-driven visual overlays. The company is cash-flow positive and actively used in production. This role will own the core 3D tracking system that aligns CAD models with physical steel parts and maintains alignment as users move. You will build upon existing internal progress and lead the system from architecture through production hardening—ideal for someone who wants to own a central technical challenge from start to shipped product.
Working at Eterio
Eterio Realities Inc. (Victoria, BC) develops Fabstation, an AR-powered manufacturing execution system for structural steel. As part of a small, experienced team, you will have direct responsibility, access to real customer environments, and a fast feedback loop from code to field use. The company is an equal opportunity employer. Candidates must be legally authorized to work in Canada.
Other
Applicants must be legally authorized to work in Canada.
