Digital Tools Training for Construction: Which Skills to Develop?

On 4 April 2016 the United Kingdom made BIM to Level 2 a contractual requirement on every centrally procured public construction project, whatever its size or value. A decade later, the training question has moved on: the hard part is no longer learning a modelling package, it is knowing what to do with the information the package produces.

From drawing software to information management

The competence that separates a digitally fluent professional from a software user is knowing which information a project owes, to whom, in what format and at what moment. That is precisely the territory the ISO 19650 series covers: an international framework for managing information with building information modelling across the whole life of an asset, split across concepts and principles, the delivery phase, the operational phase, information exchange and security.

Two things follow from that. First, the standard is deliberately tool-agnostic, so training built around it survives a change of software vendor. Second, it treats the handover to operations as part of the job rather than an afterthought, which is where most of the long-term value of building information modelling and why it matters is realised.

The four skill families worth funding

Training budgets are finite, so it helps to separate the skills that change how a project runs from the ones that only change which buttons someone presses.

Why data literacy outranks software training

A crew that can operate a platform but cannot question its output will act confidently on bad numbers. The useful skill is the sceptical one: knowing the sensor’s tolerance, the update frequency of the feed, and the difference between a measured value and a modelled estimate.

Reality capture comes with a licence attached

Drone survey is the one skill on this list that is regulated as an activity, not merely trained. In the United States, commercial drone flights over a site fall under the FAA’s Part 107 rules, which require the operator to hold a current Remote Pilot Certificate, to register the aircraft and to comply with Remote ID, with airspace authorisation needed in controlled airspace. Other jurisdictions run their own regime through their civil aviation authority, the CAA in the United Kingdom and national authorities applying the European rules across the EU. Treat this list as a pointer rather than an exhaustive account, and verify the requirements applicable where the aircraft will actually fly.

Where digital training programmes usually stall

Programmes rarely fail on enthusiasm. They fail at the point where site data has to feed something else. In its 2026 AI and Digitalization in Facilities Management Report, published in 2026 from a December 2025 survey of 760 US business leaders and 260 US facility managers at organisations of 200 employees or more, Johnson Controls found that 67% of facility managers said their organisation was already using AI, while data quality and integration issues were the biggest barrier to scaling it further.

That finding is US-scoped and self-reported, so it should not be read as a global adoption rate. It does describe a pattern worth planning around: the bottleneck sits in the information itself, which is an argument for training people to structure and check data before buying anything that consumes it. The same logic applies to how AI supports real estate planning, where the output is only ever as good as the asset records behind it.

Sequencing a twelve-month plan

1. Audit before you buy. Establish which information your projects already produce, and where it currently dies.
2. Train the coordinators first. Information management competence at coordinator level sets the rules everyone else will follow.
3. Run one pilot project end to end, including handover to whoever will operate the building.
4. Add reality capture once the model is trusted. Capturing progress against a model nobody believes in produces expensive noise.
5. Recertify on a cycle. Standards are revised and drone certificates expire, so treat this as recurring rather than one-off.

Common questions from training leads

Which digital skills matter most for construction professionals?

Information management, data literacy, reality capture and collaborative platform use. Modelling software proficiency is useful, but it is the narrowest of the group and the fastest to date.

What does ISO 19650 actually cover?

It sets out how information is specified, produced, exchanged, checked and secured across the life of an asset, across parts dealing with principles, delivery, operation, exchange and security. It does not prescribe a software package.

Is virtual reality useful in construction training?

It is used for safety induction and design review because it lets people rehearse a hazardous sequence without exposure. Maturity varies a great deal between firms, so treat vendor claims about generalised adoption with caution, and never as a substitute for the site safety training required by local regulation.

Do site managers need to learn analytics?

They need to interpret analytics, not build it. The practical target is being able to challenge a figure on a dashboard and know who to ask when it looks wrong.

How long before a digital skills programme shows a return?

That depends on project length and portfolio size, and any single figure quoted without those two variables should be treated as marketing. A defensible approach is to measure rework and information requests on one pilot project against a comparable earlier one.