Questions manufacturers ask before starting a project.

That is the first question we try to answer. MCM starts from the real product, defect, cycle time and line constraints. Where uncertainty exists, a feasibility test or pilot can be used before the full system is engineered.

Yes. Most projects are engineered around the customer process rather than selected from a fixed catalogue. Cameras, lighting, mechanics, test interfaces, software, reject logic and traceability can be adapted to the application.

In many projects the new system must communicate with existing PLCs, machines, databases or factory networks. Integration requirements are reviewed during the technical assessment so that interfaces and responsibilities are clear before build.

Support can include remote diagnostics, on-site intervention, software changes, operator guidance, maintenance planning and replacement of relevant components. The exact support model depends on the installed system and customer needs.

Yes. Depending on the application, the system can record inspection or test results, serial numbers, timestamps, recipes and other production data. The architecture is defined around the customer traceability requirements.

3D vision is useful when height, depth, volume, profile or deformation carries the decision. 2D vision is often more efficient for presence, orientation, print, colour, surface and dimensional checks that can be resolved from image contrast.

Deep learning can be useful when defect appearance varies too much for stable rule-based inspection. It still requires representative data and a realistic validation plan, so it is selected when it adds practical value rather than simply because it is fashionable.

Project duration depends on scope, mechanical complexity, purchased components, validation requirements and site coordination. A typical custom system may take several months, with milestones agreed during proposal and project planning.