{"resourceId":"uiuc-ai-weather-outage-crew-timing-2026","versions":[{"version":"external-47a87326164c30d9baac85a70e8293e59bf72d7a69c6c88fd52a86418f5ac189","resource":{"id":"uiuc-ai-weather-outage-crew-timing-2026","title":"Forecast Lead Time and the Timing of Crew Deployment for Weather-Driven Power Outages Using AI Weather Models","organization":"University of Illinois Urbana-Champaign (Golkar, Alves de Oliveira, Alipour)","sector":"Electric distribution utilities (storm restoration, mutual assistance)","geography":"Contiguous United States (county level)","publishedAt":"arXiv preprint v1, submitted 14 Oct 2026","publicationDate":"2026-10-14","eventDate":null,"sourceName":"arXiv","sourceLabel":"Academic preprint (arXiv:2610.20918)","sourceUrl":"https://arxiv.org/abs/2610.20918","evidenceClass":"academic-research","outcomeClass":"mixed","topics":["infrastructure","operating-model","governance-procurement"],"finding":"Using GraphCast and FourCastNet operational forecasts with ORNL EAGLE-I outage records (2022-2024, 2,491 counties), the authors find outage-occurrence skill (AUC-PR) falls from 0.1207 at 24h to 0.0646 at 120h, still above a 0.047 county climatology. Staging crews in the top 10% of counties captured 0.356 of high-burden outage at 24h vs 0.238 at 120h; the 72h-to-120h loss is significant. Committing at 72h with all crews beat 24h with 60% of crews; break-even availability 83.4%. A 17.5% skill gain from land-surface predictors produced no significant decision gain, joining antecedent predictors at valid time inflated skill 42-53% (data leakage), and 90% prediction intervals covered only 0.476 of tropical cyclone events.","sledRelevance":"Directly relevant to municipal utilities, public power agencies and co-ops that request mutual-aid crews before storms and are being sold AI outage-prediction tools; it gives a decision-value test rather than an accuracy claim.","evidence":"Retrospective modelling study: LightGBM hurdle model, temporal split (2022 train, 2023 calibration/selection, 2024 test), block bootstrap over 18 storm weeks, paired lead-time comparisons, SHAP attribution, conformalized quantile regression. Baselines are county climatology and a perfect-foresight oracle (0.893). Code and data on Zenodo. Not peer reviewed.","architectureImplications":"Outage models built on AI weather forecasts need a gust field (the archive lacks one; adding gust raised wind-only skill about a third), issue-time feature joins, and hazard-aware calibration. Pipelines must lag reanalysis-derived features to forecast issue time.","governanceImplications":"Procurement and validation should require decision-value metrics (captured burden, break-even crew availability) and leakage checks, not headline accuracy. Interval guarantees can fail under hazard-mix shift, so calibration evidence must be stratified.","securityPrivacyImplications":"Limited direct security relevance; uses public county-level data. Operational deployments would ingest utility OMS/AMI outage data, which carries customer and grid-sensitivity controls.","caveats":"Preprint; three years of data with a single training year; 2024 test year had an unusual hurricane season; western winter skill unexplained; county/6h resolution unsuited to feeder-level restoration; monetary values highly assumption-dependent (VoLL, restoration acceleration).","streamIds":["utility-management","emergency-services"],"roles":{"sales":"Sales framing: this is a credibility test for any outage-prediction pitch. The study shows AI weather forecasts add value over history for pre-storm crew staging, but value drops sharply beyond three days and accuracy gains do not automatically improve decisions. Use it to open conversations with public power and co-op operations leaders about how they decide when to request mutual aid, and position a bounded evaluation of their own storm history rather than promising avoided outages or guaranteed restoration savings.","engineering":"Engineering framing: treat the paper as a validation checklist. Any outage model should join land-surface features at forecast issue time, include gust information, report captured burden in staged areas, and stratify interval coverage by hazard. A proof of value can replay two or three past storms from the utility's OMS records against archived AI forecasts, comparing staging at 24, 72 and 120 hours with a climatology baseline. Feeder-level transfer is unproven because the study works at county resolution.","delivery":"Delivery framing: the operational owner is the storm-restoration or emergency operations lead, supported by analytics staff. Delivery work is a retrospective replay, a documented decision rule for mutual-aid timing, and a post-storm review loop. Acceptance should be measured on captured burden and crew-availability break-even, not model accuracy. Risks include overconfidence during hurricanes, sparse outage reporting, and staff treating a five-day forecast as actionable. Keep human commitment decisions with incident command."},"roleDetails":{"sales":{"customerProblem":"Utilities must commit costly mutual-aid crews days before storms using uncertain forecasts.","stakeholders":["Storm restoration / emergency operations manager","Public power or co-op general manager","Joint action agency or mutual-aid coordinator","Finance / rate staff"],"discoveryQuestions":["How many days ahead do you currently request mutual-aid crews, and what drives that timing?","What share of requested crews did you actually secure in recent events?","Do you retain OMS outage records by event that could support a retrospective replay?"],"valueHypothesis":"A forecast-informed staging rule may capture more outage burden than climatology at 1-3 day lead times; the study does not establish dollar savings for any specific utility.","potentialEngagement":"A retrospective storm-replay assessment of the utility's own events using archived forecasts, ending in a documented decision rule.","unsupportedClaims":["Guaranteed avoided outages (staging shortens restoration, it does not prevent outages)","Specific dollar savings for a utility","Feeder-level accuracy","Reliable five-day crew commitments"]},"engineering":{"fit":"Strong fit for pre-storm planning analytics; weak fit for feeder-level restoration dispatch.","architectureIntegration":"Combine AI weather forecast archives with OMS/EAGLE-I outage history and land-surface data in a batch scoring pipeline feeding emergency operations dashboards.","prerequisites":["Event-level outage history with consistent timestamps","Access to archived AI weather forecasts","Hazard labels for post-event review"],"constraints":["County and 6-hour resolution","Western winter skill gap unexplained","Gust field absent from the forecast archive","Single-year calibration"],"security":"Outage and customer counts from OMS/AMI should be handled under existing data-classification rules; the model itself uses public data.","proofOfValue":["Replay at least two historical storms comparing 24h, 72h and 120h staging","Run the valid-time vs issue-time correlation leakage diagnostic","Report interval coverage by hazard class"]},"delivery":{"work":"Retrospective replay, decision-rule drafting, integration of scores into storm briefings, and post-event review.","dependencies":["OMS data extracts","Forecast archive access","Emergency operations participation"],"ownership":"Storm restoration / emergency operations manager.","skillsAdoption":"Analysts need forecast-verification and leakage-testing skills; operators need training on lead-time limits and uncertainty.","governanceCheckpoints":["Leakage check sign-off before any skill claim","Incident-command approval of the decision rule","Post-storm review of staging decisions"],"acceptanceCriteria":["Proposed: captured burden at 72h exceeds climatology on replayed storms","Proposed: documented break-even crew-availability ratio for the utility","Proposed: hazard-stratified interval coverage reported"],"risks":["Overconfidence in hurricanes","Sparse or biased outage reporting","Misreading skill gains as decision gains"]}},"retrievedAt":"2026-10-29T22:06:55-05:00","enrichedAt":"2026-10-29T22:18:16-05:00","enrichmentBasis":"retrieved source","accessibilityWorkforceImplications":"Mutual-aid line crews are a constrained workforce; better timing affects crew fatigue and travel. Operators need training to interpret probabilistic outputs.","procurementImplications":"RFPs for outage-prediction tools should require decision-value metrics, leakage diagnostics and hazard-stratified calibration evidence.","operatingModelImplications":"Supports a formal, documented mutual-aid commitment rule with post-event review rather than convention-based timing.","peerReview":"preprint","sourceVerification":{"openedUrl":"https://arxiv.org/abs/2610.20918","referenceExcerpt":"Prediction intervals covered only 0.476 of tropical cyclone events against a nominal 0.90.","promptVersion":"sled-research-v4.0","model":null,"basis":"agent-reported inspection"}}}]}