The question behind retraining

If a cleaning contract brings in more technology, it is easy to say that people will be retrained. The harder question is retrained for what. A course is not a job plan. Before anyone promises a transition, the employer needs to name the work, how many people will do it and what happens to the rest of the team.

A possible new role might include setting routes, preparing equipment, responding to alerts, checking a machine after an interruption and reading the data it produces. That is an example to test, not a claim that every site will create such a role or that one person can cover every function.

What the evidence says

the International Federation of Robotics classifies cleaning robots in public places as professional service robots, a category intended for trained professional operators. Its classification does not tell us what training each employer should provide, but it does show that professional operation is part of the category rather than an afterthought.

BICSc, CSSA and the University of Surrey have launched a Future of Cleaning research initiative to examine technology, including robotics, and the training needs connected with effective deployment. Their announcement describes a research question. It does not yet establish which new jobs will exist, how many people will move into them or what pay they will receive.

Training has to fit the work

Tennant’s 2025 implementation guide describes practical support around route programming, demonstrations, training materials and ongoing use. That is useful evidence that introducing a specialised floor robot can require more than handing over a machine. It is manufacturer guidance, so it should not be treated as independent proof that every deployment is easy or that the same support is available from every supplier.

the training plan should begin with a live site and a named job. What decisions will the worker make? What problems can they solve without escalation? Who checks that the cleaning outcome meets the specification? Only then can a contractor decide whether the time, confidence and technical support are realistic.

The fairness test

A credible workforce plan should say whether learning time is paid, who pays for it, whether the role is permanent and how many openings exist. It should also say what happens if the new work needs fewer people than the tasks that changed, or if someone does not want a technical route. Those are employment and contract decisions, not questions a robot supplier can answer alone.

For a facilities manager, the practical request is simple: ask for the operating model as well as the equipment proposal. The plan should show the task, the people needed before and after introduction, the training route, human support and how the service result will be checked. That makes the promise of retraining visible enough to discuss honestly.

Read Karl Haynes’s practitioner contribution

Sources

  1. World Robotics Service Robots

    International Federation of Robotics · 2026

    Defines professional service robots as intended for trained professional operators. It is a classification and market source, not an employment study.

  2. BICSc and CSSA Collaborate on Innovative Industry Research

    BICSc · Undated, accessed 16 November 2026

    Announces research with CSSA and the University of Surrey on technology and training needs. It reports the research scope, not final findings.

  3. Robotics Implementation Guide

    Tennant Company · 2025

    Manufacturer guide describing onboarding, route programming, training and ongoing support. It is not independent outcome evidence.

Contribute to the inquiry

If your organisation has introduced cleaning technology, what skills did people actually need to operate, support and check it? Share publicly available examples of the role, the training time and what happened next.

Join the discussion on the HI company page (opens in a new tab)