How owners, contractors and government bodies are using AI to de-risk, assure and deliver rail megaprojects

Featuring speakers from Transpennine Route Upgrade, John Holland Group and the Victoria Infrastructure Delivery Authority.

One AI. Three perspectives. Real detail on what’s actually working.

Major rail projects overrun on schedule and cost by an average of 45%. The causes are well understood: optimism bias drives overconfident forecasts, availability bias pulls attention toward the most visible risks rather than the most critical ones, and strategic misrepresentation can shape plans to satisfy stakeholders rather than reflect delivery reality on the ground. Traditional methods have struggled to counter any of this, leaving most organisations reacting to problems after they've already materialised.

A growing number of rail organisations are now using AI to flip that dynamic, and this webinar is your chance to hear directly from three of them about what that actually looks like in practice. We've brought together leaders from an owner, a Tier 1 contractor, and a government delivery authority–three completely different vantage points on the same multi-billion-pound problem– each using Predictive and Generative AI on live programmes right now. Rather than talking in generalities about where AI might be headed, each speaker will go into the specifics: what their implementation actually involved, how it evolved, what it took to get it properly adopted, and the measurable difference it's making today.

In this session, you'll hear how rail delivery leaders are using AI to:

• Run AI-led risk assurance on an as-needed basis with minimal resource, keeping teams ahead of deliverability crises instead of reacting to them after the fact

• Spot risks before they become issues, drawing on a model trained on >750,000 past project schedules

• Strengthen possession assurance, giving teams the confidence to plan and protect critical track access windows without costly last-minute surprises

• Roll AI-driven risk analysis out from a single team or project to an entire contracting organisation, building real buy-in from senior delivery teams along the way

• Use conversational AI agents to interrogate schedules, run cost analysis, and generate reports and dashboards in minutes rather than days

• Give oversight and delivery authority teams clearer, faster visibility into complex programmes spread across multiple contractors and alliances

Paul Reichl - R&D Program Manager, Victorian Infrastructure Delivery Authority (VIDA)

Paul works within VIDA, delivering Victoria's ‘Big Build’: over 200 road and rail projects worth more than A$120bn, including the Level Crossing Removal Project, Melbourne Airport Rail, and North East Link. With a PhD and an R&D background, Paul is known for cutting straight to the real challenges in how contractors and delivery authorities work together. He's seen first-hand how hard it is for oversight teams to keep a handle on programmes delivered through complex alliance structures, and has championed the use of AI agents, as a way for VIDA teams to quickly understand what's happening on a project and where to focus their attention.

Paul Bradley - Group Manager, Planning, John Holland

Paul holds a central planning leadership role at John Holland, a Tier 1 contractor working across rail, transport, buildings, and energy in Australia. A former British Army engineer with experience delivering programmes in both the UK and Australia, Paul has been candid about the limitations of traditional schedule risk analysis (QSRA) and how AI-driven tools can speed it up, cut bias, and get teams to action faster. He'll talk through what changed after an early renewal project where adoption stalled at individual-project level, and how that experience pushed John Holland toward rolling AI-driven risk analysis out at an enterprise level, with a case he describes as straightforward: spot issues sooner, act earlier, and avoid the kind of costly overruns that have made headlines for the industry before.

Wes Cadby - Director of Risk Management, Transpennine Route Upgrade

Wes leads risk across TRU, a £14bn rail programme, overseeing risk processes and alliance-level risk management across the whole project. He brings experience from both rail and nuclear alongside an active role in the UK's wider risk management community. When Wes arrived, AI-driven schedule analysis was already in use across the alliances, with West leading the way, but it wasn't yet reaching his own central programme risk team or the broader programme view. Since relaunching the tool within his central team, usage has grown into embedded practice, with AI-led insight now contractualised as part of TRU's critical project planning and a routine part of possession readiness reviews. He'll talk through what it took to close that gap, the results it's delivered, and how he's now combining those learnings with what's worked in the West alliance to push adoption out further across the programme.