EV-2026-0147 · DPIA · linked to Predictive maintenance — water mains
Privacy Impact Assessment for the Westbrook CBD predictive maintenance ML model on the water main network. Concludes no PPIPA or HRIPA personal information processed; assessment closed with low residual privacy risk.
Generated from DPIA template v2026.04. Hash chained at upload. Privacy Officer e-signature attached.
None. Inputs: SCADA telemetry (asset readings, no personal data); asset register (geographic pipe metadata); historical work orders (staff-completed, no resident details); BoM weather feed. Outputs: predicted failure scores per pipe segment with GIS coords. No resident, ratepayer, or staff personal information enters the model.
Water & Infrastructure team only. Access controlled via M365 Entra group 'WaterInfra-AI-Reviewers' (8 members as of 2026-05-26). Audit log writes on each access.
Not applicable. No personal information collected, used, disclosed, or stored by this system. Confirmed against PPIPA 1998 s.4 definitions.
Low. Indirect risk only: if Council shares aggregated failure data publicly, it may inadvertently identify properties with frequent issues. Mitigation: any public release aggregated to suburb-level only, signed by Privacy Officer.
Approved by Privacy Officer 2026-05-26. Next review: 2027-05-26 or upon material change to model inputs.