Progressive Assurance: Building Confidence Throughout Delivery, Not at Handover

Why teams that defer data governance to the end pay twice, and how to avoid it.

It is Day 120 of a major project. The physical work is complete. But somewhere in a site office, a small team is frantically pulling together a handover pack: merging spreadsheets, organising photographs, hunting through email threads for missing records, and trying to decipher the handwritten notes of an engineer who left site three weeks ago.

This scene plays out across countless major programmes every year. And it really should not.

The challenge itself is not new. What has changed is the cost of getting it wrong.

Why handover matters more than ever

Regulatory frameworks like CDM and ISO 19650 now demand something very specific. They require contemporaneous audit trails. That is, records created the moment work is done, not reconstructed weeks or months afterwards. This is no longer a matter of best practice or discretion. It is a legal and contractual obligation.

Major programmes such as Crossrail and HS2 exposed the true cost of data chaos at scale. Delays, contractual disputes, rework, and reputational damage all flowed from the same root cause: information that was not captured properly the first time. Clients have learned from this. They are now writing audit-ready data requirements directly into contracts, and they expect their tier one suppliers and supply chains to deliver against them.

The TransPennine Route Upgrade set the template for what good looks like. Work delivered across 122 kilometres, 106 active worksites, and more than 400 blockades, with audit-ready handover required after every possession. The results have been significant. 

• The programme has seen a 20 percent productivity increase, with teams spending less time chasing records and more time delivering work. 
• Roughly two and a half hours have been saved per possession, time reclaimed from administration that previously sat on top of the day job. 

And, perhaps most importantly, there are no surprises at handover. Clean, compliant handovers are delivered consistently because the evidence is already there.

The problem: fragmented capture

Most projects do not have a single disciplined approach to data. Instead, they have spreadsheets sitting on a shared drive, photographs trapped on individual phones, handwritten notes in site diaries, and email threads stretching back months. By the time handover arrives, somebody is being asked to assemble a compliance-ready record from hundreds of disconnected fragments.

What we also see, even on otherwise well-run projects, is Power BI dashboards that look impressive on the surface but mask the manual effort underneath. The dashboard might be digital, but the data feeding it has taken hundreds of hours of people retyping numbers from Excel, photographing paper forms, and pulling figures out of WhatsApp threads. People think the project is digital because the output looks digital. In reality, the inefficiencies are hidden one layer down.

The result is predictable. Information is inconsistent. There is no single source of truth. Compilation at the end is manual and error-prone. Regulatory exposure increases. And payment gets delayed because the client cannot accept what has been handed over.

The solution: progressive assurance

Progressive assurance is a fundamentally different way of thinking about delivery. Rather than treating documentation as something to be done at the end of a project, every piece of data captured during delivery becomes an assurance point in its own right. Evidence is recorded in real time, structured at source, and audit-ready from the moment it is created.

Four foundational capabilities make this operational on a live programme.

1. Single Source Of Truth - All teams feed into one shared data environment. Version conflicts disappear because there are no parallel records to reconcile. When something has been signed off, everyone is looking at the same answer.
2. Real-Time Visibility - Data captured in the field flows instantly to management and to the client. Decisions are made on live information, not on yesterday's summary or last week's report. If a job is missing evidence, you know about it now, while there is still time to do something about it.
3. Audit-Ready Structure - Information is captured in formats that already meet compliance requirements. There is no post-processing step, no reformatting exercise at the end of the project. The data is fit for purpose the moment it is recorded.
4. Seamless Integration - Progressive assurance does not require ripping out existing systems. Most tier one contractors already have common data environments such as ProjectWise, Asite, or Autodesk, alongside design tools, planning software, and asset management platforms. Progressive assurance works alongside these, drawing data in from the systems that already work well and feeding structured outputs back out to the wider ecosystem. 

When these layers are connected, data flows automatically. When they are not, you end up with engineers retyping design information into quality check sheets, which is both wasteful and a source of rework when somebody makes a typo.

The strategic insight: start with three questions

Progressive assurance is not really about the tool. The technology matters, but the principles matter more. It starts with three questions answered on Day One.

1. What does compliance require?
   Regulatory and contractual data obligations need to be defined explicitly, by discipline, before construction starts. Vague intent is not enough.
2. What does the client expect to see?
   Format, frequency, and accessibility of deliverables all need to be agreed up front. If the client wants information fed back into ProjectWise as soon as something is finished, that needs to be designed in, not bolted on.
3. When, who, and how does the data flow?
   The workflow needs to be defined clearly. Who captures what, at what point in the job, using which device or template, flowing into which system.

Answer those three questions properly and the mechanics become straightforward. Ignore them and you are back in spreadsheet chaos by week six.

A continuous process of improvement

There is one further reality that nobody likes to talk about in these conversations, but it determines whether progressive assurance actually works in practice. All the data architecture and digital technology in the world will not deliver value unless the people on site use it in the right way.

The teams we have seen succeed are the ones that engage their supply chain early, explain what they are trying to achieve, and give engineers a genuine route to feed back when something feels clunky or when a form is missing a field. If your engineers do not feel heard, they will put the iPad down and go back to whatever they were doing before. It is as simple as that.

A continuous improvement loop matters. So does choosing software that is flexible enough to evolve. Rigid systems force people to work around them. Flexible systems let the process improve as the project learns.

Progressive assurance: proven on real programmes

These principles are not theoretical. They have been tested on some of the most demanding programmes in UK infrastructure.

On the Midland Main Line electrification with SPL Powerlines, the overhead line electrification scope was fully digitised. Bonding, OLE installation, daily site checks, permits, and quality assurance were all captured digitally and connected back to a central platform. Around seven million quality checks were completed digitally across the delivery period. Handover was achieved in record time because progressive assurance had already been managed throughout. When wire run testing was carried out afterwards, the scheme recorded the lowest fault levels SPL had seen on an electrification project.

On a rail signalling scheme delivered with AtkinsRéalis, nearly 300 Wheel Detection Point records were digitised using structured data capture rather than the usual manual document workflow. By capturing information once on site and structuring it for automated reporting, the project team saved an estimated 900 to 1,500 hours, improved consistency, and reduced the risk of error. More importantly, the data created during delivery became part of a reliable record that supports assurance and asset management well beyond the project itself.

On the TransPennine Route Upgrade, the productivity gains and time savings already mentioned are underpinned by structured data flowing across every discipline. Reporting takes care of itself because the data is captured in a consistent way at source. Teams are not pulled off delivery to compile reports manually.

Beyond rail and infrastructure

The same principles apply far beyond rail. Any team managing complex data, dealing with stakeholder handovers, or facing compliance requirements can benefit from progressive assurance thinking. We see it in power, utilities, civils, and the wider built environment. The sectors differ. The fundamentals do not.

The core idea is universal. Build confidence incrementally throughout delivery, rather than scrambling to manufacture it at the end.

The bottom line

The question is no longer whether progressive assurance is necessary. Across major programmes, it is becoming the expected baseline. Clients are writing it into contracts. Regulators are tightening expectations. Programmes that fail to plan for it pay twice: once during delivery, in the form of duplicated effort and manual workarounds, and again at handover, in delayed payment and strained client relationships.

The real question is whether your team gets ahead of the curve by building data governance into day-to-day operations from day one, or whether you spend the final weeks of every programme reconstructing data, chasing missing records, and managing stakeholder scepticism.

Progressive assurance is the difference between confidence and chaos. Get it right on Day One.