Every generation of engineers learns differently. Some learnt fire dynamics with a slide rule and a copy of the SFPE Handbook; others grew up with spreadsheets; many of today’s fire engineers entered the profession alongside CFD, cloud computing, and modelling tools their mentors never had. Artificial intelligence is simply the latest chapter in that story, and on its own that would be unremarkable.

What makes this chapter different is the kind of work AI can absorb. For the first time the tasks under threat are not the tedious ones at the edge of the job; they are the formative ones at its centre, the tasks a graduate has always used to learn the profession. That deserves serious discussion.

For decades, engineering organisations have balanced two competing objectives: projects need to be delivered efficiently, and young professionals need opportunities to learn, develop judgement, build relationships, and gradually take on greater responsibility. The balance has never been perfect, but it has generally worked, because the two objectives shared the same activities. A junior engineer learns by checking a travel-distance assessment against the limits in Approved Document B, by drafting the smoke-control section of a fire strategy, by chasing down the right clause in BS 9999, by building a simple zone model and watching it disagree with their intuition. Through repetition, challenge, and feedback they begin to understand not just what to do, but why.

Artificial intelligence unsettles that balance, because it is good at exactly these tasks. A report section that once took an afternoon can be drafted in seconds. Guidance can be summarised instantly. A calculation can be generated before the engineer has finished framing the question. The work still gets done. But if the machine does it, where does the learning go?

The wrong response

The instinctive response is to wall AI off from junior staff entirely. It is an understandable reflex, and it is the wrong one.

AI is becoming embedded in professional practice across every discipline, and the engineers entering the profession today will spend their entire careers working alongside it. Isolating them from these tools does not protect their development; it simply leaves them unprepared for the workplace they are about to enter, and it ignores the genuine opportunities AI creates. Used well, it can accelerate learning, expose engineers to a broader range of concepts, explain the same idea three different ways until one lands, and connect knowledge across disciplines that used to sit in separate textbooks.

The objective, then, is not to eliminate AI from professional development. It is to introduce it deliberately, in a sequence that builds competence rather than bypassing it.

Competence still has to be built

There is no shortcut to professional judgement. Engineers develop it through experience, reflection, mistakes, mentorship, and exposure to real projects, by seeing what works, what fails, and which assumptions actually matter when a design is challenged. AI cannot manufacture that.

Consider a graduate who asks a model to produce an evacuation calculation, pastes the result into a strategy, and moves on. The number may be right. But if they cannot say why the pre-movement time was chosen, what happens to the result when an exit is discounted, or whether the method even applies to the occupancy in front of them, they have not become more competent; they have only learned to transcribe. The same failure is possible with a spreadsheet, a CFD package, or a design guide used without comprehension. The tool is not the problem; the absence of understanding is.

This is why first-principles knowledge, critical thinking, and engineering judgement have to remain the foundation of professional development. AI sits on top of those foundations. It does not replace them, and a career built the other way round (fluency with the tool, no grasp of the fundamentals) is a career that cannot defend its own conclusions.

A staged approach to AI

The early stages of an engineering career are unusually formative, and they reward a deliberate sequence. In the first six months a graduate is typically still learning terminology, the regulatory framework, and the practical realities of project work. Across the first two years they build technical confidence, develop how they communicate, and start to see how engineering decisions actually get made.

Access to AI can be linked to those milestones rather than switched fully on from day one. A graduate might be expected to work a calculation by hand before reaching for an AI-assisted version, so that they can recognise when the assisted answer is wrong. They might be asked to explain a method in their own words before relying on a generated summary, or to demonstrate that they understand the assumptions behind a routine before automating it. The exact structure will differ between organisations, but the principle holds: AI should be introduced in a way that supports learning rather than short-circuiting it.

Mentoring becomes more important, not less

The most common misconception about AI is that it reduces the need for mentorship. The opposite is closer to the truth.

As AI takes over the production of first drafts and routine calculations, the centre of gravity in mentoring shifts from correcting outputs to examining reasoning. A senior engineer used to spend much of their review time fixing arithmetic and rewriting clumsy paragraphs. When the draft arrives already clean, that time is freed for the more valuable conversation, the one that develops judgement:

Why did you choose this approach? Why is this assumption appropriate for this building? What alternatives did you consider and reject? Where does this method stop being valid, and how would you defend the conclusion to a reviewer who disagrees?

Those questions were always the point of mentoring; AI just removes the busywork that used to crowd them out. The challenge for mentors is no longer to check what a junior produced, but to understand how they arrived there, and a polished AI-assisted draft can hide a shaky chain of reasoning more easily than a handwritten one ever could.

Personalised learning at scale

One of the most genuinely exciting opportunities lies in education itself. Engineers learn differently. Some learn through worked examples, others through discussion; some from a visual picture of smoke filling a compartment, others from the derivation underneath it. Historically, tailoring development to each individual has been difficult and expensive.

AI changes the economics of that. It can explain a concept several ways until one connects, generate practice problems pitched at the right level, suggest the next thing to read, and let an engineer move at a pace that suits them. None of this replaces teachers, mentors, or structured development programmes; it makes them more effective by handling the repetition and the patient re-explanation that no senior engineer has time to do at scale.

Breaking down disciplinary boundaries

Fire engineering rarely exists in isolation. A competent fire engineer has to reason across architecture, structural behaviour, human behaviour, regulation, construction, building services, operations, risk, and increasingly digital tools as well. One of AI’s real strengths is its ability to connect information across exactly these boundaries.

That makes it a natural prompt for systems thinking. A graduate working on smoke control can follow the thread into the mechanical engineering of the fans and dampers that deliver it. A question about a timber structure can open into char rates and structural fire response. A conversation about evacuation can widen into human behaviour and the management actions a strategy quietly relies on. These connections are often what separate narrow technical competence from professional judgement, and AI lowers the barrier to making them.

OpenFire as a learning surface

This is the gap we built OpenFire to close. Most engineering software answers a junior’s question by handing back a number, which trains exactly the wrong habit: answer-seeking instead of understanding. A tool meant to develop engineers has to do the opposite: show the working.

OpenFire is built around open methods, which turns a calculation from a sealed answer into something a graduate can pull apart. They can follow the derivation, see which assumptions it rests on, learn the bounds within which it applies, and change a value to watch the result respond. For a junior engineer that is the difference between being told a travel distance is acceptable and being able to trace the method behind it, the limits it assumes, and what happens to the margin when an exit is discounted. The same calculation a mentor would once have walked through on paper is there to be opened, questioned, and stress-tested, which is precisely the behaviour professional development is trying to build.

You can browse the method library and work through these calculations at the OpenFire method catalogue; it needs no account, and it is a sensible place to point a junior engineer who is ready to look behind the answer.

Building a stronger profession

The future of fire engineering will involve artificial intelligence; that much seems settled. We have argued more broadly that its place should be defined by judgement rather than automation, and professional development is where that argument is tested first. The open question is what kind of profession we choose to build around it. Used purely to cut costs and speed up delivery, AI will quietly erode the development pathways that have always produced competent engineers: the formative tasks vanish, and nothing replaces them. Introduced thoughtfully, the same technology can strengthen those pathways: personalising how people learn, widening access to knowledge, exposing engineers to a broader range of ideas, encouraging systems thinking, and freeing senior staff for the mentoring and real-project experience that actually builds judgement.

Those choices are not made by the technology; they are made by the people who lead engineering teams, a responsibility we take up in AI and leadership in fire engineering.

The objective was never to replace the next generation of fire engineers. It is to give them better tools, better support, and better opportunities to become the professionals society depends on.