AMP8 is driving significant investment in digital twins across the water sector. Ofwat's licence condition on asset management competency and the digital twin readiness problem are the same work, sequenced properly. Digital twins amplify the quality of the data they are built on, and most asset databases carry an unresolved data quality inheritance from previous AMP cycles.

The ambition

AMP8 represents up to £104 billion of allowed expenditure (including contingent allowances) across the water sector between 2025 and 2030, a 71% increase on PR19 allowances, according to Ofwat's PR24 Final Determinations. This is the largest infrastructure programme the sector has seen. Ofwat's Water Innovation Fund has £400 million available through 2030, delivered via innovation-focused competitions, including the Water Breakthrough Challenge, supporting projects across sector priorities, including digital twins, predictive maintenance, and AI-driven asset management. The technology being deployed is capable of genuinely transforming how water companies manage assets, predict failures and report performance to Ofwat.

The question that does not yet have a confident answer is this: What is the quality of the asset data these twins are being built on?

The governance insight water companies are not yet making

Ofwat's November 2025 decision to make asset management competency a licence condition, covered in our companion piece on the AMP8 regulatory shift, does not sit alongside the digital twin programme. It is the same work. The asset data governance required to demonstrate competency under a Section 13 licence condition, the statutory mechanism Ofwat uses to modify operators' licences, is the same asset data governance required to stand behind a digital twin that supports Ofwat performance reporting. The licence condition and the digital twin readiness problem are the same programme, sequenced properly.

The water companies that get this sequence right will have both ISO 55001 competency and digital twins they can defend. Those that run them in parallel, or twin-first, will resolve the data governance question twice, at significantly higher cost.

What a digital twin actually requires

A digital twin is not a visualisation tool. At its core, it is a dynamic, continuously updated model of physical assets and their behaviour. It is only as accurate as the data feeding it.

In an asset-intensive infrastructure environment, that data comes from multiple sources: CMMS records, CAFM systems, IoT sensors, inspection reports, condition assessments, maintenance histories, and design documentation. For the twin to support genuine operational decisions, predictive maintenance, resilience modelling, and Ofwat performance reporting, all of those sources need to be governed.

Governed means: accurate, complete, consistently formatted, validated at source, and maintained over time.

In most water sector organisations, the honest assessment of those data sources is more complicated.

The asset data inheritance from previous AMP cycles

Legacy CMMS and CAFM implementations have years of accumulated data quality issues: duplicate asset records, inconsistent condition-assessment methodologies, missing attributes, and maintenance histories logged against the wrong asset. These are the predictable consequences of deploying technology fast, across AMP6 and AMP7, before data governance standards were mature enough to keep up.

Most water companies entering AMP8 carry this inheritance. It has rarely been fully resolved because the business case for a comprehensive data remediation programme is harder to make than for a new capability. And so organisations move from one platform upgrade to the next, migrating the same data quality problems each time.

When that data becomes the direct input to a digital twin expected to support Ofwat performance claims and operational decision-making, the unresolved problems become material.

The amplification problem

Digital twins, like all AI and analytics platforms, amplify the quality of the data they process. This is not a technical nuance. It is the central risk of any digital twin deployment in an environment where the underlying data is ungoverned.

Feed a validated, well-governed asset dataset into a digital twin platform, and you get accurate, trustworthy operational insight. Feed it ungoverned, duplicated, inconsistently captured data, and you get confident outputs from unreliable inputs.

The confidence is the problem.

A digital twin that presents a misleading infrastructure health picture with apparent precision is more dangerous than no model at all. It creates the illusion of certainty where uncertainty exists.

In an AMP8 environment where Ofwat holds water companies to account on asset resilience, leakage performance, and infrastructure health, the stakes of that misplaced confidence are not theoretical. Each outcome underpinned by twin outputs is directly exposed under the ODI regime, which charged the sector £157.6m in net underperformance for 2023-24 alone. A misleading twin translates into penalty exposure and defensibility gaps the finance director cannot hedge.

The Chief Digital Officer reads amplification as a build-quality problem. The Asset Management Director reads it as an evidence problem. The Regulatory Liaison reads it as an Ofwat exposure problem. All three are right. The fix sits upstream of all three, in the governance of the asset data the twin is built on.

The governance path, and the next ninety days

ISO 55001 is the management system standard that provides the governance architecture for delivering asset data to the quality level required for digital twins to operate. Implementing it means establishing how asset data is created, validated, maintained, and reviewed across the whole lifecycle, not just at the point of migration or deployment. It creates the audit trail that lets an organisation stand behind the data its digital twin is drawing from.

Brainwave Asset Intelligence's ADDR framework (Assess, Design, Deliver, Realise) delivers that path for infrastructure organisations. The Assess phase identifies the specific data quality gaps relative to the planned digital twin deployment, including duplicate records, missing attributes, inconsistent condition classifications, and unvalidated sensor metadata. Design builds the remediation programme. Deliver and Realise take the organisation from its current state to governed, twin-ready asset data.

For digital twin programme leads, the practical diagnostic sits in three steps:

  1. Audit the asset data sources feeding the planned twin. Classify each by governance maturity: ungoverned, partially governed, fully governed. The distribution tells the board whether the twin is ready to be built on or whether the sequence is wrong.
  2. Run ADDR Assess against the specific twin scope, not the whole asset estate. A targeted gap map scoped to the twin delivers a decision-grade answer without rescoping the whole asset estate.
  3. Sequence data remediation before the build, not in parallel. Parallel remediation against a live twin burns the twin's credibility before it has a chance to earn it, and it is the pattern Ofwat will see first when scrutiny arrives.

This work must precede twin deployment, not follow it. A post-deployment data remediation programme against a live twin is significantly more complex and more expensive than resolving the data quality problems before the model is built. More importantly, it means the twin has been producing outputs, potentially shaping operational decisions and Ofwat submissions, on the basis of data that was known to be ungoverned.

The continuous layer: Data Governance as a Service

Water data does not become defensible once and stay defensible. Networks expand, assets age, supply chains rotate, platforms upgrade. A digital twin built on a 2026 baseline begins drifting the moment the team that built it moves on.

Data Governance as a Service (DGaaS) is Brainwave Asset Intelligence's cross-sector model for keeping the asset record defensible between reporting cycles rather than at them. The full treatment, applied consistently across asset-intensive sectors, sits in the DGaaS anchor.

The sector observation

The ambition across the sector is well-directed. The technology being deployed is capable. What determines whether that investment produces the intended outcomes, accurate operational insight, reliable Ofwat performance data, trustworthy AI-generated recommendations, is the quality of the governance work that precedes it.

New Civil Engineer's analysis from April 2026 identifies the end of water company self-reporting as imminent, with independent automated monitoring replacing operator-reported data. The governance work required for that transition, verified, auditable, asset-level data, is identical to the governance work that makes a digital twin trustworthy. The two agendas are not separate. They are the same work.

The water companies that get the sequence right will have digital twins they can stand behind. Those that reverse the sequence will find out the hard way what their data was worth.

Key takeaways

  • Digital twins amplify data quality: accurate, governed data produces reliable insight; ungoverned data produces confident failures
  • Most water sector asset databases carry an unresolved data quality inheritance from previous AMP cycles
  • In an AMP8 environment with Ofwat performance accountability, a digital twin built on poor data is a regulatory risk, not just a technical problem. ODI exposure was £157.6m sector-wide in 2023-24 alone
  • The ISO 55001 licence condition and the digital twin readiness problem are the same work. Run them as one programme, sequenced properly
  • ISO 55001-aligned data governance must precede digital twin investment. The sequence determines the outcome

Sources and further reading