Lall, ShrutiPillay, Nelishia2026-04-092026-04-092026-01S. Lall and N. Pillay, "Anticipatory Edge Intelligence: A Foundational Enabler for Resilient Critical Infrastructure Systems" in IEEE Internet Computing, vol. 30, no. 01, pp. 71-80, Jan.-Feb. 2026, doi: 10.1109/MIC.2026.3657339.1089-7801 (print)1941-0131 (online)10.1109/MIC.2026.3657339http://hdl.handle.net/2263/109487Critical infrastructure systems (CISs), such as power grids, water networks, and transportation systems, operate under stringent requirements for timeliness, resilience, and coordinated response. As these systems become increasingly data-driven and automated, decisions must often be made under uncertainty and with limited tolerance for delay. This position article advocates for anticipatory edge intelligence (AEI) as a conceptual framing for designing edge-enabled intelligence in CISs, with resilience and containment as primary objectives. AEI emphasizes the generation, exchange, and operationalization of short-horizon anticipatory information at the edge to enable coordinated, preemptive action before degradation propagates. The article examines key challenges faced by CISs, identifies opportunities where anticipatory coordination can enhance system-level resilience, and uses an illustrative scenario to motivate this perspective. By articulating AEI as a research and design lens, this work aims to guide future investigation into resilient, edge-enabled infrastructure systems.en© 2026 IEEE. All rights reserved, including rights for text and data mining, and training of artificial intelligence and similar technologies.Edge AIArtificial intelligence (AI)ResilienceDegradationDelaysCritical infrastructureCloud computingReal-time systemsDecision makingComputer architectureEdge computingCritical infrastructure systems (CISs)Anticipatory edge intelligence (AEI)ModernityTransport systemInternet Of Things (IoT)Power gridTraffic congestionTight couplingWater networkPre-emptive actionControl systemDisasterForecastingPhysical systemOnline learningNoisy dataLearning mechanismsSmart gridGraph neural networksCascading failuresEdge nodesDeep uncertaintyDigital twinStatic systemMulti agent reinforcement learningCyber physical systemsInterdependent componentsDisaster scenariosFlight pathPostprint Article