{"resourceId":"oati-uamps-basin-electric-ai-agent-etag-transmission-challenge-2026","versions":[{"version":"external-7e932d4bae7c3053a4149360ca31669fca15668c51f34b00cc05a950b980de0f","resource":{"id":"oati-uamps-basin-electric-ai-agent-etag-transmission-challenge-2026","title":"UAMPS and Basin Electric win OATI AI use-case challenge for an agent that chooses book-out vs physical flow, buys transmission and builds e-Tags","organization":"OATI (Open Access Technology International) with Utah Associated Municipal Power Systems (UAMPS) and Basin Electric Power Cooperative","sector":"Public power joint action agency and G&T electric cooperative: wholesale scheduling and transmission","geography":"Intermountain West and Upper Midwest, United States","publishedAt":"2026-11-10","publicationDate":"2026-11-10","eventDate":null,"sourceName":"OATI news","sourceLabel":"Vendor announcement of workshop competition winners","sourceUrl":"https://www.oati.com/news/ai-use-case-challenge-winners/","evidenceClass":"vendor-claim","outcomeClass":"emerging","topics":["developers-agents","infrastructure","operating-model"],"finding":"At OATI Energy Conference 2026, UAMPS (wholesale supplier to 50 community-owned systems in seven states) and Basin Electric (G&T owned by 138 member co-ops in nine states) jointly won an AI use-case workshop with an agent concept: decide whether a transaction should be a financial book-out or physical flow under current market conditions, find and buy transmission on the right path, build the e-Tag, and pre-build an alternative route and tag in case of curtailment. OATI will develop it with schedulers and traders on its OASIS and e-Tagging systems; each organisation gets a year of complimentary service. OATI's guidance lists constraints that make pilots 'rise or stall': NERC CIP, operator trust and data availability.","sledRelevance":"A concrete agentic-AI use case owned by a municipal joint action agency and a co-op, areas thin in recent editions. It is a concept, not a deployment, but shows where public power sees agent value: frequent, time-pressured, data-heavy scheduling decisions.","evidence":"Vendor announcement; no prototype, accuracy or time-saving data. Winning idea will be developed; complimentary service creates a commercial relationship.","architectureImplications":"Agent operates on existing OASIS reservation and e-Tag workflows; actions include purchasing transmission, which implies transactional authority and audit requirements.","governanceImplications":"An agent that buys transmission and submits tags needs explicit authority limits, human approval thresholds and market-compliance review.","securityPrivacyImplications":"Bulk electric system scheduling sits near NERC CIP scope; agent credentials and actions require strong access control and logging.","caveats":"Concept stage; vendor marketing; no independent assessment; results may not apply to distribution-only municipal utilities.","streamIds":["utility-management"],"roles":{"sales":"Sales framing: joint action agencies and G&T co-ops make the same scheduling and transmission decisions many times a day under deadlines. UAMPS and Basin Electric chose that workflow for an AI agent, which signals where public power sees real value. This is a concept announcement, so the bounded offer is a use-case scoping and controls design, not an outcome promise: define the decision, data, authority limits and success measures before any build.","engineering":"Engineering framing: the agent would sit on OASIS, transmission purchase and e-Tag workflows, so it needs reliable market and availability data, transactional APIs and full action logs. Treat purchases and tag submissions as high-consequence actions requiring thresholds and human approval at first. Map NERC CIP applicability early, isolate agent credentials, and design curtailment fallbacks that operators can override. Nothing here is proven yet; test in shadow mode against scheduler decisions.","delivery":"Delivery framing: start with working sessions to document how schedulers decide today, then run the agent in shadow mode and compare its choices and timing with human decisions before granting any purchasing authority. Scheduling managers own outcomes; compliance owns NERC CIP review. Proposed acceptance criteria are agreement rates with schedulers, tags built before deadlines, and zero unauthorised transactions. Risks are operator distrust, data gaps and dependence on one vendor's platform."},"roleDetails":{"sales":{"customerProblem":"Schedulers repeatedly connect market, transmission and tagging steps under deadline pressure, and redo them when curtailments hit.","stakeholders":["Scheduling and trading manager","Compliance (NERC CIP) lead","Member utility representatives","IT/OT security"],"discoveryQuestions":["How many book-out vs physical-flow decisions are made per day?","How often do curtailments force re-tagging?","Which actions could an agent take without human approval?","What data feeds are reliable enough for automation?"],"valueHypothesis":"An agent with bounded authority could reduce time spent assembling transmission reservations and tags, measured in shadow mode before use.","potentialEngagement":"Use-case scoping and shadow-mode evaluation design for an agency's scheduling workflow.","unsupportedClaims":["That the agent exists or works today","Any time or cost savings","That it meets NERC CIP requirements"]},"engineering":{"fit":"Joint action agencies, G&T co-ops and municipal utilities that schedule wholesale power and buy transmission.","architectureIntegration":"Agent layered on OASIS, transmission purchase and e-Tag systems with action logging and approval workflow.","prerequisites":["Documented decision rules","Market and ATC data feeds","Approval thresholds"],"constraints":["NERC CIP scope","Market rules on tagging deadlines","Vendor-platform dependency"],"security":"Separate agent credentials, least privilege for purchases, immutable logs and operator override.","proofOfValue":["Shadow-mode agreement with schedulers","Time-to-tag comparison","Curtailment fallback test"]},"delivery":{"work":"Map current decisions, define authority limits, run shadow mode, then stage limited authority.","dependencies":["Vendor development","Scheduler time","Compliance review"],"ownership":"Scheduling and trading manager.","skillsAdoption":"Schedulers trained to supervise and override agent actions.","governanceCheckpoints":["NERC CIP applicability review","Authority-limit approval","Go-live decision on shadow results"],"acceptanceCriteria":["Agreed agreement rate with human decisions","All tags submitted before deadlines in test","Zero unauthorised transactions"],"risks":["Operator distrust","Bad data driving purchases","Lock-in after free year"]}},"retrievedAt":"2026-11-10T21:56:37.357111-06:00","enrichedAt":"2026-11-10T22:03:27.938369-06:00","enrichmentBasis":"retrieved source","accessibilityWorkforceImplications":"Schedulers and traders shape the agent; their expertise and trust are prerequisites.","procurementImplications":"Free first-year service may lead to lock-in; agencies should define exit and data-ownership terms.","operatingModelImplications":"Shifts schedulers from executing steps to supervising an agent and handling exceptions such as curtailments.","peerReview":"unknown","sourceVerification":{"openedUrl":"https://www.oati.com/news/ai-use-case-challenge-winners/","referenceExcerpt":"Pilots rise or stall on NERC CIP, operator trust and data availability.","promptVersion":"sled-research-v4.0","model":null,"basis":"agent-reported inspection"}}}]}