WWC Research Note · September 23, 2026 · 5 min read
Why Human-in-the-Loop Design Matters in Wastewater Decision Support
Read-only architecture, operator investigation and auditability — and why autonomous control is outside the current scope.
The operator remains responsible
Wastewater treatment is critical infrastructure. Operators carry responsibility for process stability, effluent quality and safety. Any software that participates in their decisions should strengthen that responsibility, not obscure it.
Waste Water Care is designed as decision support: it organises information, highlights risk and records actions. It does not act on the plant.
A read-only operating boundary
The current MVP scope includes no PLC writes, no chemical dosing, no blower actuation and no control-system commands. Future integrations are planned as read-only ingestion from historians or SCADA, subject to security review.
A read-only boundary simplifies safety reasoning. The worst outcome of a wrong forecast is a misleading view — which the confidence and data-quality layers are designed to reduce — not an incorrect command sent to equipment.
Supporting investigation, not replacing it
Good decision support helps operators ask better questions: Is this probe healthy? Has influent load changed? What did we do last time? The product surfaces data quality, drivers and evidence so that investigation is faster and better documented.
Auditability
Every alert in the workflow moves through explicit states, and each transition creates an audit event: who acknowledged it, who was assigned, what was observed, what action was recorded, and why it was closed.
Reports draw on these records. Where AI assistance drafts text, it is labelled as an AI-assisted draft requiring human review, and approval is an explicit human step.
Why not autonomous control?
Autonomous optimisation is sometimes presented as the natural endpoint of AI in treatment plants. It is not part of the Waste Water Care scope. Any move in that direction would require validated models, extensive operational testing and regulatory engagement that have not taken place.
Takeaway
Human-in-the-loop is not a limitation to be engineered away. In this domain it is the design principle that makes the rest of the product trustworthy.
See these ideas in the interactive MVP.