Beyond the buzz: using AI in construction

AI is entering every sector, but construction lags behind. The industry is complex, fragmented, and slower to adopt digital tools. Even BIM, which has existed for decades, is still seen by some as an unnecessary expense.

That hesitation comes at a price. From 2000 to 2022, global construction productivity improved only 10%, just one-fifth the rate of the overall economy. The gap shows how much opportunity remains untapped.

The question isn’t about whether AI will matter, but how to use it effectively.

Getting the focus right

Everyone loves to use the term “AI Strategy”, but AI is not a strategy. It’s a tool. And like any tool, its impact depends on how precisely it’s used.

Are we solving for cost control? Improving productivity? Reducing risk? Bridging skill gaps? Focus is key. Without a clear objective, companies spread resources thin and see little return. The process is simple:

Define the goal → Pilot → Measure → Scale.

At Ascentis, we’ve learned that small, targeted wins inspire far more confidence than sweeping declarations. It’s not about ticking the “AI-ready” box. It’s about solving one real problem at a time. Case in point: At Ascentis, we use AI to match CVs to job descriptions, saving hours of manual screening while improving our hiring rate. Simple and effective.

LLM vs SLM

While Large Language Models (LLMs) like ChatGPT can support broad tasks like writing reports or drafting specifications, the real value may come from Specialised Language Models (SLMs) trained on our data: contracts, BIM models, schedules, and cost plans. With either of these options, there’s no need to aim for the moon. The construction supply chain is so wide and information so fragmented that AI can facilitate incremental changes at the working level in many aspects.

Getting the most out of AI

1. You can’t optimise what you haven’t digitised.

AI can’t work with data that doesn’t exist. Far too many contractors still rely on paper drawings, handwritten logs, and fragmented spreadsheets. The first step is structured data capture with consistent naming, tagging, and digital workflows.

Even modest improvements, such as digital daily logs or centralised drawing repositories, lay the groundwork for more intelligent systems later.

2. You’re only as good as your data

Every project leaves behind valuable information: schedules, cost reports, RFIs, and design revisions. Unfortunately, this data is useless if spread across PDFs, emails, spreadsheets, and WhatsApp message threads. Data that’s riddled with errors, unstructured or locked in silos can’t be fed to AI systems to process properly. If we invest the time to clean, structure, and unify this legacy data, we could have a goldmine for forecasting, planning, and risk prediction.

We need to evolve from data storage to data intelligence using what we already have to make better decisions.

AI + BIM

BIM is a natural starting point for AI integration, elevating it from a coordination tool into a decision-making engine.

Beyond detecting clashes, it can:

• Automate quantity take-offs.
• Simulate 4D construction sequences.
• Provide real-time cost and schedule feedback.
• Suggest design or procurement alternatives for better value.

For example, AI-enhanced BIM can flag design elements that increase embodied carbon or identify procurement bottlenecks long before construction starts.
Used this way, AI doesn’t just prevent errors, it predicts performance.

The benefits

AI is changing how project teams work, not by replacing people, but by removing friction.
Much of project management today is administrative: gathering progress updates, comparing drawings, or verifying invoices. These are precisely the kinds of repetitive, rule-based tasks AI can handle faster and more accurately.

That means less time chasing information and more time making decisions that actually move the project forward.
When used well, AI becomes an invisible assistant — helping teams plan, coordinate, and troubleshoot in real time.

1. Faster, better decisions: AI tools can synthesise information across schedules, cost plans, and correspondence to surface trends a human might miss, like early signs of slippage, budget drift, or design inconsistencies. The result is more proactive management. Teams can correct course before issues turn into disputes or delays.
2. Democratizing access to expertise: AI makes explicit knowledge (brand standards, engineering rules, contract clauses, and even lessons from past projects) widely accessible. It’s a leveller for junior engineers and managers who may not have decades of experience. The best firms will use AI to accelerate learning, not replace judgment.
3. Enhancing collaboration: By integrating with BIM and scheduling platforms, AI allows designers, contractors, and owners to view the same live data, breaking down silos that have long slowed the industry. Everyone works from one version of the truth, which means fewer misunderstandings and rework.
4. Elevating the role of project managers: As AI takes over routine coordination, what it will not be able to replace is the implicit/tacit knowledge around know-how and years of experience anticipating risk, managing relationships, and making trade-offs between time, cost, and quality. This is the core of what PM expertise is, shifting from administration to strategic leadership.

The pitfalls…and the opportunities

With every software vendor now marketing “AI capability,” it’s easy for developers to get lost in the noise. What can AI actually do for a project? Which tools integrate best? How do you measure success?

Project Managers will play a central role in answering these questions. As AI influences design, cost, and scheduling decisions, we’ll need frameworks for data integrity, traceability, and ethics to ensure accountability remains human.

This will require significant upskilling.

Tomorrow’s PM must understand not just construction but also digital ecosystems, effectively becoming an orchestrator of both people and data, evolving the profession from coordination to digital orchestration.

Final takeaway: AI won’t replace people, but it will reward those who use it wisely.

AI’s potential in construction is real but uneven. The winners won’t be those who rush to adopt it everywhere, but those who apply it with focus and discipline, guided by leadership that values both human and machine insight.

At Ascentis, we see AI not as a shortcut, but as a tool that strengthens judgment, improves collaboration, and enhances project performance.

AI won’t replace people, but it will reward those who use it wisely.