AI and Fire Safety — Where We Are Now

Artificial intelligence is beginning to make its presence felt in fire safety engineering, though — as in most of the built environment professions — the pace of change is slower than the headlines suggest. The fundamentally technical and judgement-heavy nature of fire engineering means that AI tools are currently most useful as aids to human experts rather than replacements for them. But the trajectory is clear, and understanding where AI is heading in fire safety is important for anyone involved in the sector.

AI in Fire Modelling

Computational Fire Dynamics (CFD) modelling has been a staple of performance-based fire engineering for decades. Machine learning techniques are now beginning to accelerate fire modelling — reducing the computational time required for complex simulations, enabling more scenarios to be tested, and improving the calibration of models against experimental data. AI-assisted fire modelling will make performance-based fire engineering faster and more accessible on projects where it would previously have been impractical.

AI-Assisted Compliance Checking

One of the most promising near-term applications of AI in fire safety is automated compliance checking — using AI to review building drawings against fire safety requirements and identify potential issues. Several technology companies are developing platforms that can parse architectural drawings, extract geometric information, and check against Approved Document B requirements for travel distances, compartment sizes and escape route widths. These tools will not replace the judgement of a chartered fire engineer, but they may significantly accelerate the review process and catch obvious issues earlier.

AI and the Golden Thread

The Building Safety Act 2022's golden thread requirement — the obligation to maintain a comprehensive digital record of building information throughout the building's life — creates a structured data environment that AI is well suited to exploit. AI tools that can monitor the golden thread, flag when maintenance records are falling behind, identify changes to the building that may affect safety, and support the building safety manager in their ongoing duties are a natural development.

What AI Cannot Do

The core of fire engineering is judgement — the ability to understand a complex building, identify the specific risks it presents, and devise engineering solutions that address those risks within the constraints of the design. This requires deep technical knowledge, professional experience and the ability to reason about novel situations. These are capabilities that current AI systems do not possess, and the Building Safety Regulator's focus on human competence means that the professional accountability for fire engineering decisions will remain with chartered engineers for the foreseeable future.