AMP8 needs a 50 percent bigger water workforce: can induction and reality capture take the strain off the people you already have?

AMP8, the water industry's 2025 to 2030 investment period, will need roughly a 50 per cent larger water workforce, a figure Water UK projects for the sector to deliver everything it has committed over the 2025 to 2030 cycle. Recruitment cannot fully close that gap on the timeline the work demands, so the lever that is actually in your control is throughput: how quickly a person who has already turned up gets safe enough to be productive. That is where online induction and a 3D capture of the site people can see before they visit earn their place, and it is a different problem from hiring.

How big is the AMP8 skills gap, really?

Start with the numbers, because the framing depends on getting them right. Within water specifically, Water UK projects a 50 per cent workforce increase to deliver the full AMP8 programme.

The often-quoted figure of 312,300 additional workers is broader. The Energy and Utility Skills workforce demand estimates put more than 312,300 additional skilled workers across the whole energy and utilities sector, that is gas, power, water and waste, by the end of 2030. It is not a water-only number, and it is worth being precise about that when you are planning against it.

The pressure is not only about hiring more people. EUSR, the utilities sector's skills registration scheme, notes that delivery depends on people, skills and competence, not capital alone, with more than 30,000 additional skilled roles needed across England and Wales during AMP8, spanning engineering and operations. Layered on top is an ageing profile and a retention problem, with a large share of the existing workforce approaching retirement and many engineers looking elsewhere.

Why is recruitment the wrong place to spend all your energy?

Because the maths does not resolve fast enough. Even with strong apprenticeship and recruitment pipelines, the people who close a 50 per cent gap are not fully deployable on day one, and the retirement cliff removes experience at the same time.

So the question a water delivery lead should be asking is not only how many people can we bring in, but how much productive capacity can we get from the people and contractors we already have. Every hour a competent subcontractor spends stood in a site office waiting to be inducted is capacity you paid for and did not use. Across hundreds of dispersed water sites, treatment works, pumping stations, reservoirs, network sites, that lost time compounds quickly.

What is a site induction actually for, and why is it not onboarding?

This distinction matters and it gets blurred constantly. Onboarding is a relationship with an employer: contracts, systems, culture, the things a permanent hire needs once. A site induction is about a specific place: its hazards, its access routes, its emergency arrangements, the live plant you must not walk into.

On an AMP8 programme the same subcontractor might visit twenty different water sites in a year. Each is a distinct environment with distinct risks. The induction has to travel with the site, not the person, and it has to be repeatable at scale without a supervisor delivering it in person every time. That is a different job from onboarding, and conflating the two is how programmes end up delivering neither well.

It is also worth being honest about where a generic PDF induction falls down. We have written before about the hidden cost of PDF site inductions: people click through without reading, comprehension is unproven, and a document cannot show someone what a specific site actually looks like before they stand in it.

How does reality capture change the induction itself?

A conventional online induction can move the paperwork and the safety video out of the way before the visit, so the person arrives having already completed the generic content and only needs a short site-specific orientation. That alone recovers time, and it is why the online induction category exists.

Reality capture (recording a real site with drones, 360 cameras or scanners) adds the part a document cannot: the actual site. Using drone and 360 imagery, and where the detail warrants it a full 3D scan, you can build an induction that walks a subcontractor through the real environment they are about to enter. They see the access point, the wash-down area, the chemical dosing room, the confined space they must not approach, before they ever set foot on site.

For a dispersed water estate this is the meaningful shift. Sending a person to build a bespoke induction for every one of hundreds of remote sites is not realistic. Capturing a site once, then delivering a photoreal induction for it through a platform your teams and your supply chain already use, is. The capture is the input; the induction people actually complete is the product.

Where is this the wrong answer?

It is not a fit everywhere, and pretending otherwise would waste your money. A construction site that changes its layout and hazards week by week will outrun a fixed capture, so there the honest advice is to keep the site-specific element live and lightweight rather than scan a state that is obsolete in a fortnight.

Reality capture also does not replace competence assurance. It gets someone safely oriented to a place faster; it does not certify that they can do the task. EUSR registration, SHEA cards and role competence still sit where they always did. Induction is the gate to the site, not a substitute for the ticket to the trade.

And a captured induction is only as current as the site it was captured from. A works part-way through an AMP8 upgrade will change, so plan a regular schedule for capturing it again rather than treating a scan as permanent. Selling it as a one-and-done would be the wrong promise.

The practical read for AMP8

The workforce shortfall is real and the recruitment work has to happen. But the gain you can bank this year is throughput, not headcount. Getting a competent subcontractor from arrival to safe productivity faster, consistently, across a dispersed estate, is capacity you already own and are currently leaving on the table.

Online induction gets the generic content done before people arrive. Reality capture makes the site-specific part genuinely useful across hundreds of locations without sending someone to each one. If the same sites are being inducted into over and over, which describes most of a water estate, that is exactly the shape of problem this solves. Where languages are a factor across a mixed supply chain, the same captured induction can also carry one induction across every language so comprehension does not depend on the reader's first language.