Job Description

This role owns the full perception pipeline for an embedded robotics platform operating in real-world construction environments. The successful candidate will integrate AI-based object recognition models — covering people, equipment, and obstacles — into real-time embedded systems with deterministic, resource-constrained runtime requirements. The position sits at the intersection of embedded systems engineering and applied AI, requiring both deep hardware-awareness and practical deployment experience. This is not an AI research or cloud-side ML role; the focus is on building perception systems that work reliably on the device, in the field.


Job Requirements

Requirement

  • Design and implement full perception pipelines for embedded robotics platforms
  • Integrate AI-based object recognition models into real-time embedded systems
  • Architect memory-efficient, deterministic runtime systems for edge deployment
  • Optimise inference performance under resource constraints (memory, compute, thermal) across heterogeneous embedded accelerators
  • Integrate multi-sensor inputs including camera, LiDAR, and IMU
  • Connect perception outputs reliably to control and safety systems
  • Analyse latency, memory usage, and numerical stability across the pipeline
  • Ensure robust operation across diverse edge cases and real-world field environments

Qualifications

  • Hands-on embedded software development experience in modern C/C++
  • Deep understanding of real-time systems, scheduling, and performance optimisation
  • Proven experience deploying AI/ML models in edge or embedded environments
  • Development experience in resource-constrained embedded computing environments
  • Working knowledge of computer vision fundamentals: camera models, projection, coordinate transformations
  • Familiarity with model export and deployment pipelines (e.g. ONNX) and quantisation concepts
  • Ability to navigate accuracy/latency/hardware trade-off decisions in production

Preferred

  • AI model optimisation experience with heterogeneous embedded accelerators (DSP, NPU, or equivalent)
  • Robotics or autonomous systems development background
  • LiDAR processing or basic sensor fusion experience
  • Safety-critical or deterministic system design experience