Why I think AI belongs in the modern construction toolbox Shelley Copsey, CEO and co-founder of FYLD, says AI might not wear a hard hat, but it belongs firmly on the construction site. Written by Shelley Copsey Published on 10 July 2025 Our experts We are a team of writers, experimenters and researchers providing you with the best advice with zero bias or partiality. Written and reviewed by: Shelley Copsey Direct to your inbox Sign up to the Startups Weekly Newsletter Stay informed on the top business stories with Startups.co.uk’s weekly email newsletter SUBSCRIBE A few years ago, I sat in a room with senior leaders from one of the UK’s largest utilities. We were reviewing a safety incident — thankfully not fatal — but serious enough to prompt a formal investigation. A hazard had been missed. A site crew had carried on, unaware. A supervisor hadn’t had enough information to intervene. “We’ve got systems everywhere,” one executive said to me, “but I still feel like I’m managing in the dark.” That moment has stayed with me, because I’ve heard versions of it again and again. When I founded FYLD, it wasn’t to build another piece of fieldwork software. It was to fix a deeper problem I’d seen across infrastructure, utilities, and construction for years: the people doing the hardest highest-risk work were operating without real-time support – and those trying to lead them were doing so with limited visibility and delayed data. The more time I spend in these conversations, the clearer it becomes: so many of the risks we’re talking about today in construction can be solved. But it takes a shift in mindset; one that prioritises visibility, accountability, and agility, not digitisation for its own sake. It means using technology like AI, not as a trend, but as a practical way to empower teams, prevent problems, and make faster, better decisions. After two decades working across infrastructure, operations, and data, here are six of the most common risks I see every day that could be prevented with better visibility — and what we’ve learned about solving them at FYLD. 1. Managers are running sites from the rearview mirrorConstruction work is fast-moving and unpredictable. But decisions are often made based on what happened yesterday. Or, worse, what someone remembers over the phone. That’s where AI and the right technology can fundamentally change how decisions are made. For our customers using FYLD, instead of waiting on delayed updates or relying on memory, managers can see the actual conditions on site as the job begins. A short video captured by the crew could be analysed instantly, flagging missing safety controls or unusual risks, and giving supervisors the information they need to step in, guide, or approve in real time. 2. Communication hasn’t kept up with the way work happens nowHybrid and remote work are now part of the fabric of construction, but most teams are still trying to manage with disconnected tools. Crews use phone calls, WhatsApp, spreadsheets, emails. It works… until it doesn’t. AI and modern platforms can close these gaps by embedding communication into the work itself. Instead of chasing down updates, teams collaborate on a single timeline where photos, messages, permits, and blockers all live. AI can flag when something’s off and automate alerts. The right person gets notified at the right moment—with the context to act. That’s how you reduce noise, align faster, and keep work moving. 3. Safety is still too reactiveToo often, safety is treated as a compliance task rather than a core driver of operational performance. Risk assessments are filled out to satisfy legislative requirements, not to drive smarter behaviour. And more importantly, the data that does get collected is rarely used to inform decisions in real time. This is where video and AI can shift safety from a box-ticking exercise to a live risk management tool. When a site walk is captured on video, AI can assess whether key hazards are being considered, whether control measures are visible, and whether the job should move forward. It can prompt additional actions or approvals. 4. We talk about labour shortages, but not untapped potentialThere’s no question the sector is facing pressure on workforce capacity. But what we’re seeing on the ground is that it’s not just about hiring and recruitment, it’s about making better use of the people you already have. With the right tools, every job becomes a training opportunity. AI can surface lessons from previous work, highlight recurring risks, and match newer staff with just-in-time coaching. Workflows can embed expertise into the flow of delivery, without relying on someone being physically present. This not only improves quality and consistency, but it also allows you to scale without increasing headcount. 5. Supply chain partners are being managed with guessworkFor project owners and lead contractors, one of the hardest things is understanding what’s actually happening on site – especially when the work is being delivered by subcontractors using different tools, reporting styles, and levels of digital maturity. Modern field execution tools can standardise how work is documented, tracked, and assessed, regardless of the owner. AI scans for patterns across jobs and contractors, spotting quality issues early and flagging risk trends before they spread. 6. Most projects are still too slow to respond when things changeDisruption is inevitable in construction, whether it’s weather delays, late materials, workforce constraints, or shifting regulations. Most don’t have the visibility or flexibility to adapt quickly. By the time leaders realise something’s gone off track, the damage is done. Real-time intelligence changes that. By using AI to process job data as it’s generated – photos, videos, site notes, permit activity – you can spot blockers immediately and respond in hours, not days. This kind of responsiveness isn’t just a nice-to-have. It’s how you improve margins, protect timelines, and keep trust with stakeholders. Where we go from hereAI is a long way from replacing skilled people in this industry. But it can give them sharper tools, clearer context, and faster feedback. It can turn supervision into collaboration. Compliance into foresight. And it can help leaders lead with clarity, not assumption. If we want to unlock safer, faster, more scalable field operations, we need to stop working in hindsight and start connecting in real time. I built FYLD to help the construction and infrastructure industries lead from the front, not the rearview mirror. By Shelley Copsey, CEO and co-founder of FYLD FYLD is the AI-powered fieldwork intelligence platform helping infrastructure and construction teams work safer and smarter. FYLD was named to the UK Startups 100 in 2025 for its impact on operational performance, real-time decision-making, and frontline productivity. Learn more about FYLD Share this post facebook twitter linkedin Tags News and Features Written by: Shelley Copsey