Perception Engineer

NOX METALSDetroit, MI
Onsite

About The Position

Perception Engineer | Nox Metals | Detroit, MI American factories deserve a supply chain that moves as fast as they do. The next generation of American manufacturing is being built right now. Nox Metals is the supply chain behind it. Nox Metals is a technology company in Detroit supplying aluminum plate, bar, and rounds to aerospace and defense manufacturers. We use software and automation to supply metal to American factories faster than the industry thought possible. We need a Perception Engineer to give the factory eyes, ears, and a sense of touch. Vision systems on every machine. Vibration sensors on every saw. Thermal sensors on every cut. Every signal turned into data, every dataset turned into a better cost model, a smarter nest, a tighter cycle, and a more reliable cell. Nox Metals is an employer with real opportunity for long term career growth, this is a place to build a career, not just hold a job.

Requirements

  • Strong technical foundation in computer vision, signal processing, and sensor systems, judged on capability not years
  • Hands-on with industrial cameras, machine vision platforms, and 2D and 3D imaging
  • Experienced with non-vision sensors, accelerometers, vibration, thermal, current, force, and the analog and digital interfaces that connect them
  • Fluent in modern ML and DL frameworks (PyTorch, JAX, OpenCV, or comparable) and the surrounding tooling
  • Comfortable shipping the full pipeline, you collect the data, you train the model, you deploy it on edge or cloud, you monitor it, and you retrain it
  • Experienced integrating perception into PLCs, robotics, and production software systems
  • Comfortable on the software side, you write the integration, hit the API, structure the payload, reason about the model in production
  • Tightly partnered with ML, automation, and controls teams in past roles, comfortable writing specs, filing tickets, and shaping the tools you use
  • Cracked with AI tools, modern software, and data, you build perception systems that talk to the rest of the company
  • A builder at heart, you would rather ship a working v1 today than ship a perfect v3 next month
  • High attention to quality, every model, every dataset, every sensor calibration
  • Always thinking about how to make systems better, you do not accept "this is how we have always done it"
  • Absolutely customer obsessed, every signal you capture eventually shows up as a part on a customer's dock
  • User obsessed, you sit with the operator, you watch them run the machine, and you build perception they actually trust
  • A team player with a good attitude, you make the floor better for everyone around you
  • Someone who takes ownership, if it senses, sees, or hears, it is your responsibility
  • Precise under pressure and reliable
  • Organized and detail oriented
  • Committed to safety, you follow every protocol, wear your PPE, lock out tag out, and never cut corners that put people at risk
  • High agency, you handle big items alone and ask for help when needed
  • Low ego, you walk the floor, you crawl into cells, you do the unglamorous work because it needs to get done
  • Able to lift up to 50 lbs and operate heavy equipment on your feet when the floor needs you
  • Not afraid to work outside normal hours when America demands it
  • Never says "that's not my job"

Nice To Haves

  • Experience deploying perception systems in metals, machining, or industrial environments
  • Background with vibration analysis, predictive maintenance, or condition monitoring
  • Experience with thermal imaging or thermography in process control
  • Hands-on with edge inference platforms, NVIDIA Jetson, OpenVINO, Coral, or comparable
  • Background in 3D vision, structured light, stereo, or LiDAR
  • Open source work, published research, or a portfolio of perception systems you have built

Responsibilities

  • Own the perception stack across the entire factory, vision, vibration, thermal, force, and any other sensor that gives us a signal worth capturing
  • Build vision systems for robotics, inspection, dimensional verification, defect detection, and material identification
  • Build sensor systems on saws, CNC equipment, and automated cells, vibration for blade wear, thermal for cut conditions, current draw for load, and whatever else makes the cell smarter
  • Feed perception data into job costing, cycle time prediction, and nesting optimization, the perception layer is what makes WAYNE and our cost models actually accurate
  • Own the full pipeline end to end, sensor selection, mounting, calibration, data collection, labeling, training, deployment, edge inference, monitoring, retraining
  • Build edge inference where it makes sense and cloud inference where it makes sense, you make the call
  • Build the data pipelines that capture every frame, every waveform, every reading, the data that powers our ML and operational intelligence
  • Partner tightly with the ML, automation, controls, and software teams, you sit at the seam where hardware, signals, models, and software all meet
  • Run experiments, ablations, and real-world tests to prove what works on the floor, not just on a benchmark
  • Always ask questions, never guess when something is unclear
  • Look at every sensor, every model, every signal, and figure out how to make it better
  • Work safely every shift and hold your teammates to the same standard
© 2026 Teal Labs, Inc
Privacy PolicyTerms of Service