Problem Definition

Status: Draft · Part 1 of the solution specification · June 2026

Built from the canonical strategy document (../README.md). Anchors the solution definition and requirements that follow.

Protagonist

An Australian regional or mid-sized water utility / council water business — the tier below the metropolitan majors (Sydney Water, Melbourne’s retailers, SA Water, Perth’s Water Corporation, the South-East Queensland utilities). It operates a brownfield network (mains, treatment, pumps, valves) largely laid mid-century, and it has neither the capacity nor the budget for a multi-year, big-vendor digital-twin program.

Who this actually is — named examples. This is a large, well-defined tier, typically serving tens of thousands to a few hundred thousand connections:

• Victorian regional urban water corporations — e.g. Barwon Water (Geelong), Coliban Water (Bendigo), Goulburn Valley Water (Shepparton), North East Water (Wodonga), East Gippsland Water (~35,000 people), Wannon Water (Warrnambool), Central Highlands Water (Ballarat).
• NSW council-run Local Water Utilities (LWUs) — 89 of them, serving more than 1.8 million people outside the Sydney Water and Hunter Water areas — e.g. Central Coast Council, Tamworth Regional Council, Riverina Water (Wagga / Eastern Riverina), MidCoast Council.
• South-East Queensland (the initial go-to-market focus — see below) — a layered structure: Seqwater, the state-owned bulk-water authority running the SEQ Water Grid, sells treated bulk water to the distributor-retailers and council water businesses that distribute and retail it: Urban Utilities (Brisbane, Ipswich, Scenic Rim, Somerset, Lockyer Valley), Unitywater (Moreton Bay, Sunshine Coast, Noosa), and the council-run water businesses City of Gold Coast (one of the country’s largest council water businesses, serving the Gold Coast region), Logan and Redland City Council.
• Other Queensland regional providers — council utilities such as Townsville, Cairns and Bundaberg.
• TasWater — Tasmania’s single state utility (~221,000 water connections, ~11,400 km of mains): a clean upper-”mid” example of the target profile.

Hunter Water and Icon Water (ACT) sit just above this tier — useful reference points, but larger state-owned corporations rather than the core beachhead. (NSW DCCEEW — Local Water Utilities, Victorian regional water corporations, Seqwater — SEQ structure, Business Queensland — SEQ providers, TasWater)

Initial go-to-market focus: South-East Queensland. Of this tier, near-term access is warmest in SEQ — so the first conversations target the SEQ providers, beginning with the council-run water businesses (notably City of Gold Coast) and the distributor-retailers (Urban Utilities, Unitywater). The broader tier above is the expansion path once the SEQ beachhead is proven.

flowchart TB
    SW["Seqwater<br/>state bulk-water authority"]:::bulk
    subgraph DR["Distributor-retailers"]
        UU["Urban Utilities ★"]:::focus
        UNI["Unitywater ★"]:::focus
    end
    subgraph CWB["Council water businesses"]
        GC["City of Gold Coast ★"]:::focus
        LOG["Logan"]:::council
        RED["Redland"]:::council
    end
    SW --> DR
    SW --> CWB
    subgraph LEGEND["Legend"]
        direction LR
        lb["Bulk authority"]:::bulk
        lf["★ GTM focus"]:::focus
        lc["Other"]:::council
    end
    classDef bulk fill:#dbeafe,stroke:#2563eb,color:#1e3a5f;
    classDef focus fill:#dcfce7,stroke:#16a34a,color:#14532d;
    classDef council fill:#fef3c7,stroke:#d97706,color:#7c2d12;

South-East Queensland: Seqwater supplies bulk water to the distributor-retailers and the council water businesses below it; first conversations target the ★ entities.

The core problem

It cannot trust what it knows about the assets it already operates.

The same asset — a buried asbestos-cement (AC) main, a pump, a valve — is described across GIS, the asset-management system, document/EDRMS records, SCADA, contractor and consultant surveys, and paper as-builts. These sources disagree, are incomplete, and carry no record of which is right or where they came from. Two surveys of the same network can place it differently; an install date or material may be recorded in one system and contradicted in another; large stretches are simply unknown.

Compounding this, most existing tools identify each asset in a single “language” — usually the engineering tag — and treat the more common operational names as second-class. But a water operator thinks in street names, asset nicknames, and locations, not engineering tags. When the only first-class key is the tag, the people actually running the network are locked out of finding their own assets, and records that share a real-world thing under a different name never get connected.

Why it bites now

• Aging assets are failing measurably — a national median of >12.7 water-main breaks per 100 km per year, with much of the network past design life — and the sector is being told to move to risk-based, data-led renewal.
• Risk-based management is only as trustworthy as the data beneath it, and the data isn’t trustworthy. You cannot prioritise renewal, or safely plan an AC-pipe intervention (service-continuity and asbestos-exposure risk), on records that contradict each other and don’t flag their own uncertainty.
• Regulatory expectations are tightening (see the call to action below) — the demand to prove you know your assets is moving from good practice toward obligation.

The consequence of the gap

Unplanned main breaks and service outages; unsafe or over-cautious AC-pipe works; boil-water and compliance exposure; emergency reactive cost instead of planned renewal; and capital spent in the wrong place because the risk model was fed bad data. The remediation cost of wrong asset data, once you are operating on it, dwarfs the cost of getting it right.

The regulatory direction — a reason to act now

The bar for proving you know your assets is rising, and it is shifting from good practice toward obligation. This is the “tackle it now” call to action: the cost of building a trusted asset picture is far lower when done deliberately and ahead of a deadline than when scrambled together to satisfy a regulator later.

• In Australia, the pressure is already converging. Economic regulators (IPART in NSW, the ESC in Victoria) tie price determinations to demonstrated asset-management performance; the Bureau of Meteorology’s National Performance Report benchmarks every utility publicly; NSW Health governs drinking-water safety; and the WSAA is advocating for risk-based asbestos-cement management. Independently, each asks the same question: can you show, with evidence, that you understand your assets’ condition, criticality and risk?
• The leading indicator is overseas — and regulation in water travels. In the UK, Ofwat is making asset-management maturity an explicit licence condition: water companies must demonstrate they understand asset condition, criticality and performance, and prove it through ISO certification or independent third-party assessment. What is a UK licence requirement today is a credible template for Australian expectations tomorrow. (Ofwat — asset-management maturity proposal, Water Magazine)

A utility that builds a trusted, provenance-true asset picture now — one that can show what it knows, what it doesn’t, and on whose authority — gets ahead of this curve cheaply and on its own terms. Acting early is the opportunity; waiting turns it into a compliance scramble.

Supporting cases (public)

Three public, citable cases for use in lead conversations. Each is annotated honestly with how directly it maps to the reconciliation problem — overclaiming would be the wrong way to open a conversation with this buyer.

1. Australia — “Quality Level D” asset records and underground service strikes (directly on-thesis). Under the national standard AS5488, much of Australia’s buried-utility location data sits at Quality Level D — the lowest tier, derived from existing records alone with no field verification. As the locating industry puts it plainly: “asset records show where an asset was meant to go, not where it ended up after the install crew moved it around a rock or a tree root twenty years ago,” and Before-You- Dig information is explicitly indicative only. The consequence is not abstract: underground service strikes are “among the most frequent and most catastrophic excavation incidents in Australia” — striking a pressurised water, gas or HV service can kill. The risk recurs in practice: in July 2024, roughly 100 Wollongong (NSW) properties lost water or pressure after a contractor struck a main during roadworks — a discrete instance of this systemic exposure. (The public reporting illustrates the frequency and impact; it does not, on its own, attribute that strike to records.) This is the thesis in one case: contradictory, unverified, source-blind records turn routine work into safety and service risk. (Smart Scan Locators — utility strikes in NSW, BYDA Best Practice Guide, Illawarra Mercury — Wollongong main strike)

2. Havelock North, New Zealand, 2016 — a small supplier’s catastrophe (root cause was source water; the Inquiry substantiates an information → response/rectification-time link). A campylobacter outbreak in a small reticulated supply made more than 5,000 people ill, with 45 hospitalised and a possible contribution to four deaths. Its value to us is that the Government Inquiry documents, on the public record, how fragmented and outdated information lengthened both the response and the rectification:

• Slower response. The District Council had no emergency response plan, no draft boil-water notice and no up-to-date contact lists for vulnerable people, schools and childcare. The Inquiry found that a pre-prepared boil-water notice would have reached consumers about two hours earlier than it did.
• Incomplete / outdated asset and catchment knowledge. The catchment held numerous private bores, “some known, some not,” offering direct contamination pathways; the Te Mata aquifer was incorrectly believed to be confined and the bores secure; and the District Council “did not properly manage plant and equipment maintenance or keep records of that work.”
• Slower rectification. Restoring a safe supply stretched for months — bore 3 was not reactivated into public supply until March 2017, roughly six months after the August 2016 outbreak.

Honest framing: the root cause was contaminated source water (sheep faeces after heavy rain), not asset-data reconciliation — but the Inquiry directly substantiates the thesis that missing, outdated and unrecorded information amplifies harm and slows recovery, in exactly our buyer tier (a small, regional supplier). (Government Inquiry — Part 1 Overview, DIA — Stage 2 report, AWA — lessons from Havelock North, NZ Herald — failures date back years)

3. England, 2013 — a contamination event traced to an asset missing from the record (discrete and regulator-documented; an adjacent kind of records gap). In October 2013, a private supply serving a farm and several properties (including a holiday let) returned water positive for E. coli and coliform bacteria. The Drinking Water Inspectorate found the cause was a private spring supply that was completely absent from the local authority’s register despite being the primary water source — so the back-pressure contamination pathway into the mains went unmanaged. A second DWI case (Pickering, North Yorkshire, Nov 2014) is similar: a meter found running backwards revealed an undetected illegal cross-connection to an untreated borehole. Honest framing: this is a records gap of a specific kind — an asset missing from a register — rather than the contradictory-across-systems network records at the centre of our thesis, and it is UK. But it is a discrete, regulator-documented case in which an incomplete asset record directly enabled a contamination event in exactly the small-supply context our buyer operates. (DWI Case Study 2013/13, DWI Case Study 2014/11)

Supporting detail: research on water-main failure investigation finds utilities often cannot establish provenance after the fact — “it may not be clear who performed the work or if a work order was submitted… there may not be evidence that work was performed at all” — the exact provenance gap STREAM’s source-tagging is built to close. (ASCE, Journal of Performance of Constructed Facilities)

Why existing tools don’t close it

• Extraction / AI tools (e.g. DIGATEX) digitise the document graveyard once — they do not keep live, changing sources reconciled, and AI-extraction is itself commoditising.
• Class libraries (e.g. Datum360, now Autodesk) define what data should exist — they do not reconcile what does exist across contradictory sources at runtime.
• Enterprise platforms (AVEVA, Cognite, Bentley iTwin) are heavyweight, enterprise-priced, and largely resolve conflicts by last-writer-wins or manual stewardship — the wrong shape and the wrong scale for this buyer.
• GIS / CMMS are each a single system of record — they are one of the contradicting sources, not the thing that reconciles them.
• Tag-centric information models privilege one identifier (the engineering tag) as the master and demote every other name. They cannot serve an operations audience that navigates by common names, and they miss merges that only a non-tag identifier would reveal.

The unmet need

No one offers this buyer continuous, reversible, provenance-true reconciliation — a layer that:

• ingests every source and merges them by scoped identity, where the engineering tag, the operational common name, the GIS feature ID, and the SCADA point are all co-equal, first-class keys;
• retains conflicts with their provenance rather than silently overwriting them;
• un-merges automatically when a source changes or is withdrawn;
• and produces one trusted picture that says plainly what we know, what we don’t, and who said so.