Industrial robots for the engineering lab.

A university buys for two reasons at once, and they pull in opposite directions. Teaching wants equipment that survives a cohort and matches what industry runs. Research wants a platform open enough to take apart and instrument.

Most departments want two of these three and discover they needed the third. The teaching lab and the research platform are different purchases with different budgets, and the staff who will run both are usually nobody's line item.

Another 378 universities and institutes worldwide run DOBOT robots in published research, from Imperial College London to KAUST. The full list is available on request.

The reason to publish this figure rather than a testimonial is that it is checkable. Every one of these institutions put its name to a paper, not to a quote in a brochure.

Institution marks identify published research and teaching use of DOBOT robots. They do not imply endorsement of Zeta42 or DOBOT.

For teaching, the argument is continuity. Students who learn on the arms that factories actually run graduate employable, and a department teaching on one family from first year to final project is not retraining itself every September.

For research, the argument is access. A platform that hides its control stack is a demonstration rather than an instrument, and a final-year project that cannot reach below the interface is a project about the interface.

• Cobot programming, vision and force control for engineering and computer-science degrees.
• Final-year projects and research on humanoids and data-collection rigs.

In practice faculty come first. A department that has never run a cobot cannot supervise a project on one, so the useful sequence is almost always staff, then curriculum, then students.

Operator and drag-to-teach workshops. AI Foundations and AI at Work for faculty and admin teams. Short courses through career centres and student clubs.

Technical and vocational colleges usually want something different again: employability rather than novelty, and a qualification an employer recognises on sight. That case is made better on the industry page, which is written for the people who will be doing the hiring.

A campus project is slower than a factory one and the order matters more, because a department that takes delivery before its staff are trained has bought a store cupboard.

Set up to serve models to every bench over the local network, for thirty people querying at once. Model library, datasets and notebooks loaded and backed up. The model, the weights and the reason it slows down are all in the room, which is the point for a department that teaches how the thing works.