Decades of Operational Data. Almost None of It Informing Decisions Today.
The average water utility holds more than 30 years of telemetry, inspection records, risk assessments, and maintenance history. It lives in databases no one has time to query, in systems that do not talk to each other, and in the heads of senior engineers who are approaching retirement.
The result is capital programme decisions made on incomplete information, CCTV inspection budgets deployed without risk ranking, and reactive maintenance spend that costs three to ten times more than planned intervention.
"The AI assistant does the same calculations that would take me two days and digging around in the database to find."
Network responsible, Norwegian municipal water utility
From Asset Risk Score to Ranked Rehabilitation Plan in Under 60 Seconds
Delta is an AI decision support layer built exclusively for water and wastewater utilities. It connects your live SCADA telemetry, graph-based network model, PipeFusion asset risk scores, and rehabilitation cost data into a single reasoning engine that answers operational questions in plain language.
Ask about a pressure zone, a failing asset, a catchment with elevated infiltration, or the cost consequence of a planned intervention. Delta returns a reasoned, auditable answer grounded in your actual network data — not generic hydraulic knowledge drawn from the internet.
No calibrated hydraulic model required. Delta delivers flow estimates and fault context using graph algorithms across pipe topology — removing the single biggest barrier to AI adoption in water utilities. You can be operational in weeks, not years.
Generic AI Knows About Water Networks. Delta Knows About Yours.
Every procurement team asks this question. The answer is not about interface or convenience — it is about data, governance, and accountability. A general-purpose AI assistant has no knowledge of your pipes, your risk scores, or your SCADA alarms. It cannot generate a ranked rehabilitation list for your network because it has never seen your network.
Delta is not a general AI configured for water. It is a decision support system built from the architecture outward with water utility data, water utility governance requirements, and water utility accountability at the centre.
| ChatGPT / Copilot | Delta | |
|---|---|---|
| Data source | General internet knowledge, public documents | Your SCADA telemetry, asset risk scores, pipe topology, and rehabilitation cost data |
| Network knowledge | Generic hydraulics textbooks and public standards | Graph-based model of your actual network, built on your topology and flow history |
| Rehabilitation ranking | Not possible | Yes — cost-ranked, with transparent methodology traceable to named data inputs |
| SCADA alarm context | Not possible | Yes — upstream topology, adjacent risk scores, and recent telemetry in one response |
| Data leaves your environment | Potentially yes, depending on configuration | Never — stateless inference, EU-hosted, no training on customer data |
| Auditability | Reasoning is opaque | Full — every recommendation traces back to named risk scores, cost inputs, and network parameters |
| NIS2 compliance | Requires significant configuration and legal review | Designed in from the architecture up, not retrofitted |
| Calibrated hydraulic model needed | Not applicable | No — graph-based estimation removes this barrier entirely |
| Domain scope | Everything — with no genuine depth in any sector | Deliberately bounded to water and wastewater network decisions |
"We went through the solution with our IT department, and they were impressed by how InfoTiles handled data security. Everything runs within Microsoft, no data leaves the system, and nothing is used for model training."
Utility manager, Norwegian water utility
What Delta Costs You. Far Less Than Not Using It.
A mid-scale utility managing 300 to 600 km of network can expect measurable productivity recovery, CCTV inspection cost avoidance, and reactive maintenance reduction within the first 12 months of deployment.
Manual rehabilitation prioritisation for a 500 km network takes two to four days of senior engineer time per cycle. Delta returns the same output in under 60 seconds. At a loaded rate of €800 per day, four cycles per year recovers €6,400 to €12,800 per engineer.
CCTV inspection costs €8 to €25 per metre. Delta-driven prioritisation typically enables a 20 to 30 percent reduction in wasted inspection spend — representing €100,000 to €150,000 in annual cost avoidance for a typical mid-scale utility.
Unplanned pipe failures cost three to ten times more than planned interventions. Advancing one high-risk segment per year avoids emergency repair costs of €30,000 to €150,000 per incident.
Regulatory risk reduction. A single notice of non-compliance can carry direct costs of €50,000 to €500,000. Utilities that demonstrate a transparent, data-driven rehabilitation methodology to regulators also reduce the risk of capital programme challenge and allowance reduction.
Four Data Layers. One Reasoning Engine. Zero New Infrastructure Required.
Answers are composed from multiple data layers your InfoTiles deployment already holds. No new sensors, no new data collection, and no hydraulic model calibration required.
"The LLM is the reasoning layer. The network model is the knowledge base. Neither is sufficient without the other."
Predictive Maintenance, Rehabilitation Prioritisation, and I&I Analysis
Delta's scope is deliberately bounded to areas where InfoTiles data provides genuine depth. Capabilities outside this scope are not surfaced.
Query risk scores, explain the drivers behind a specific segment's ranking, and compare candidate assets for intervention sequencing.
Generate cost-ranked intervention lists from risk data and rehabilitation method costs, with transparent methodology and full audit trail.
Surface hydraulic context around an active alarm — upstream topology, adjacent risk scores, recent telemetry — without manual correlation.
Combine SewerIntelligence I&I telemetry with PipeFusion risk scores to prioritise CCTV inspection plans, structured and ranked.
Decision Support in Action: I&I Inspection Prioritisation Using Live Network Data
The example below is an actual prompt run against a utility network. Delta combines SewerIntelligence I&I telemetry with PipeFusion risk scores to generate a prioritised CCTV inspection plan — structured, ranked, and ready to act on.
Live With Norwegian Water Utilities. Here Is What Their Teams Report.
"We went through the solution with our IT department, and they were impressed by how InfoTiles handled data security. Everything runs within Microsoft, no data leaves the system, and nothing is used for model training."
"The AI assistant does the same calculations that would take me two days and digging around in the database to find."
"If you are wondering about something, you can actually ask follow-up questions."
"It will find the same thing I find, just much faster."
How Delta Works: Your Network Data Stays Inside Your Boundary. Always.
Delta uses a retrieval-augmented generation (RAG) approach. When you ask a question, the relevant segments of your network model, risk scores, and reference documentation are retrieved from your own data store and supplied to the language model as context for that specific query. The language model does not hold your data between sessions, is not trained on your data, and is not aware of any other utility's network.
The language model provides reasoning capability. Your network model provides the knowledge base. Neither is sufficient without the other — which is precisely why a general-purpose AI tool cannot replicate what Delta does.
Human-in-the-loop is embedded at the formal recommendation layer. Delta surfaces evidence and methodology. Decisions requiring sign-off remain with the operator or asset manager. This is an architectural design principle — it means Delta's outputs are defensible in capital programme review, regulatory audit, and procurement evaluation.
Infrastructure: Microsoft Azure AI Foundry, EU-hosted, stateless inference, no training on customer data. The language model is replaceable; the retrieval pipeline is where the domain specificity lives.
EU-Hosted, NIS2-Aligned, and Built for Critical Infrastructure Procurement
Delta is Norwegian-built and operated. For utilities managing critical national infrastructure, the provenance and governance of AI tooling matters structurally — not just technically.
Runs on Microsoft Azure AI Foundry within EU/EEA regions. No data is processed outside EU/EEA.
Each session is fully independent. Nothing persists beyond the session boundary. No customer data is used for model training by any party in the pipeline.
Designed to operate within NIS2 and applicable Norwegian security legislation. Access controls and retrieval pipelines were built to satisfy procurement requirements from the start.
Authentication and access control integrate with your existing enterprise identity management and conditional access policies.
Operational in Weeks. Live With Norwegian Utilities. International Rollout Underway.
Delta is in production with Norwegian water and wastewater utilities. Expansion to UK, Netherlands, and further European markets is underway.
Deployment requires no new sensors, no calibrated hydraulic model, and no rip-and-replace of existing systems. Delta deploys on top of the network intelligence your InfoTiles platform already holds. A typical utility is operational within weeks of contract signature.
For procurement teams. The System Description, Data Processing Agreement template, Service Level Policy, and Security Reporting Policy are available at the links above. If you require additional documentation for your procurement process, contact us directly.
Why Delta — The Mathematics of Change, the Hydrology of Convergence
In the Heimskringla — Snorre Sturlason's thirteenth-century chronicle of the Norse kings, the land is described through its landmarks: the headlands, the rivers, and the varder. Stone cairns built on high ground to mark passage, to say someone was here, this is the way. Each cairn assembled by human hands, one stone at a time, accumulating into a permanent signal visible across distance and time.
Delta carries three layers of meaning in a single symbol. In mathematics, it is the symbol for change, the difference that matters, the anomaly that demands attention. In hydrology, a delta is where flows converge and are understood, where a network meets the wider world. And the cairn: the ancient waymarker built stone by stone by those who came before, so that those who follow know where they are and where to go next.
Your network already holds the knowledge — in telemetry, in risk scores, in decades of operational data. Delta assembles it, makes it speakable, and passes it forward — so the next person always knows where they stand.
See Delta running against your network data.