Machine vision engineers and optical test specialists

Machine vision engineers assess whether the required defects or measurements are visible under realistic production conditions. Request a feasibility prototype before committing to inspection hardware or a detection approach.

Work covers camera and lens selection, lighting trials, image calibration and acquisition software using interfaces such as GigE Vision or USB3 Vision. Engineers compare representative good and defective samples, document acquisition constraints and define acceptance criteria. They identify where part presentation or optical changes are needed before further software development.

Industrial inspection specialists develop automated optical inspection software for component verification and surface-defect detection. They turn product quality requirements into measurable decisions and validate them against production images.

• Rule-based inspection: Develop dimensional checks and pattern matching in OpenCV or HALCON, with tolerances linked to acceptance criteria.
• Deep learning inspection: Train defect detection or segmentation models with PyTorch and integrate inference through ONNX Runtime where appropriate. Evaluate missed defects and false rejects on held-out samples, including difficult product variants.

Automated optical test specialists write software for stations that measure optical performance or verify assemblies through image-based tests. They translate test specifications into repeatable sequences with explicit pass/fail limits.

C++ or Python development can connect cameras and measurement instruments, control test execution and record results against product serial numbers. Specialists implement calibration checks and measurement uncertainty assessments where required. Test software can retain images alongside measurements so quality engineers can investigate failures and compare results across station revisions.

Connect inspection software to production controls and maintain it as products or station conditions change. Machine vision software engineers investigate intermittent failures and extend existing applications without discarding validated inspection logic.

• Line integration: Implement PLC handshakes through OPC UA or vendor protocols, coordinate image triggers and return inspection decisions within the station's timing constraints.
• Ongoing support: Diagnose acquisition faults and processing delays. Add product recipes, regression-test changes against recorded datasets and maintain versioned configurations so releases remain traceable.