Condition-based maintenance for the grid's most critical assets
Unified DGA, partial discharge, and thermal data into survival-style risk scores for substations and lines.
Calendar-based maintenance treats every asset the same; condition data reveals which ones actually need attention.
The Challenge
A regional transmission operator managed 1,200+ substation assets and 4,000 miles of line with condition data scattered across dissolved gas analysis (DGA) labs, partial discharge monitors, thermal imaging reports, and CMMS work history. Maintenance was calendar-driven; critical failures still caused unplanned outages averaging 14 hours per event. Planners lacked a unified risk view, so capital was allocated by age and replacement cost rather than actual condition.
The Innovoco Solution
We unified DGA, partial discharge, thermal, and work history into health scores with survival-style risk models that estimate remaining useful life and recommend maintenance actions—integrated into CMMS and crew scheduling with planner overrides and audit-ready rationale.
Phase 1 — Data unification and health index
Integrated DGA lab results, online partial discharge monitors, thermal imaging, nameplate data, and 10 years of CMMS work orders into a normalized asset health index. Validated against known failure modes and recent outage investigations with reliability engineering.
Phase 2 — Risk scoring and maintenance optimization
Trained survival models on censored failure data to estimate probability of failure over 1-, 3-, and 5-year horizons. Risk scores drive prioritized maintenance recommendations published to CMMS with cost-benefit rationale for each intervention.
Key implementations
Multi-source condition fusion
Normalized scoring across DGA, partial discharge, thermal, and visual inspection data with configurable weights by asset class and vintage.
Survival-style risk models
Probability-of-failure estimates over multiple horizons, handling censored data (assets that have not yet failed) correctly—unlike naive classification approaches.
CMMS and crew integration
Maintenance recommendations publish directly to work management with priority, estimated cost, and supporting condition evidence for planner review.
Regulatory evidence packs
Exportable asset health reports with data lineage and model assumptions for rate case filings, NERC compliance, and public utility commission inquiries.
Threshold alerting
Real-time alerts when DGA or partial discharge readings cross configurable thresholds, with escalation paths to control room and asset management.
Technical Innovation
Survival models handle the censored-data problem inherent in asset health: most assets have not yet failed, so standard classification yields biased estimates. By modeling time-to-event with covariates, the system produces calibrated risk curves that maintenance planners and regulators can interrogate.
Impact
- 22% reduction in emergency trips and unplanned outages on monitored assets.
- 11% shift in O&M spend from reactive to proactive maintenance.
- 100% asset coverage with unified health scores across all substation and line assets.
- Regulatory evidence preparation reduced from weeks to hours with exportable condition reports.
Maintenance planners allocate capital based on actual asset condition—not age or calendar cycles—and regulators see defensible evidence that reliability investment is targeted where it matters most.
Explore this outcome on your stack
We map scope, guardrails, and rollout to your data boundaries and teams—practical next steps, not a generic slide deck.
60 min · Free · No obligation
More case studies
Explore adjacent patterns from the same production playbook.
AI-Powered Business Intelligence
Ask questions in plain language, get governed answers with charts — no analysts, no ticket queue.
Read more
Customer support & voice
Triage, sentiment-aware routing, CRM-grounded answers at scale.
Read more
Regulated onboarding (KYC / AML)
Document intelligence, exception queues, immutable evidence for auditors.
Read more