Perception Engineer

Ovii's Interpretation of the Role

Origin is hiring a Perception Engineer to build 3‑D perception pipelines that enable autonomous construction robots to understand and navigate complex sites. The role blends computer‑vision research, ROS 2 development, and edge‑AI optimization to deliver production‑ready perception stacks.

Role Snapshot

• Develop 3D perception components
• Build ROS 2 nodes for sensor data
• Train deep‑learning vision models
• Fuse camera, LiDAR and inertial data
• Optimize models for edge deployment
• Curate real and synthetic datasets
• Collaborate across robotics teams

Must-Have Requirements

• Python
• C++
• ROS 2
• Deep‑learning (PyTorch or TensorFlow)
• Point‑cloud processing (PCL, Open3D)
• Linear algebra, probability and geometry
• ROS 2 development
• point‑cloud processing
• deep‑learning vision model training

Nice-to-Have Signals

• Camera‑LiDAR calibration
• Real‑time optimization libraries (Ceres, GTSAM)
• camera‑LiDAR calibration
• real‑time optimization libraries

Work Setup

• Location: Bengaluru, India
• Work mode: ONSITE
• Employment type: Full-Time

Not Specified in JD

• Visa sponsorship
• Salary range
• Remote eligibility
• Education requirement
• Certifications
• Relocation
• Travel
• Security clearance
• Coding test
• Portfolio
• GitHub
• Writing sample
• Cover letter

What You'll Likely Work On

• Design and deploy 3‑D perception modules using depth sensors, LiDAR and RGB cameras
• Create ROS 2 nodes that transform point clouds, depth maps and image streams into environment models for planning
• Train and integrate deep‑learning models for semantic segmentation and surface analysis
• Engineer robust sensor‑fusion pipelines that combine visual, inertial and spatial data for localization
• Benchmark and tune perception models for NVIDIA Jetson edge platforms with TensorRT/ONNX
• Collect, synthesize and manage high‑quality datasets; automate training and experiment tracking
• Partner with manipulation, navigation and cloud teams to ship production‑ready perception stacks

Good Fit If You Have

• Experience with camera‑LiDAR calibration is a plus
• Familiarity with real‑time optimization libraries such as Ceres or GTSAM
• Strong academic or project background in computer‑vision or robotics perception