2.1 Planning Constrained by volatility of demand, inflexible planning cycles and fragmented visibility, today’s planning approaches struggle to keep pace with rapid change, leaving factories slow to respond. The future is intelligent and resilient planning – an evolution towards AI and simulation-first adaptive systems that learn continuously, anticipate disruptions and synchronize decisions across the value chain. Autonomous and federated multi-agent planning Evolutions of themes AI agents enable self-healing operations by rebalancing operational signals (e.g. production, assets, workforce, energy and demand) within factories, while a federated network of factories securely co- plans in real time, absorbing shocks and adapting to changing demand.3 Schedule adherence On time in fullSimulation-driven, reactive — Simulation-based planning provides cross-functional visibility. — Operational signals are used to adjust schedules and resources during disruptions, with actions remaining largely human-led. — Federated planning pilots connect OEMs with Tier-1 and Tier-2 suppliers through shared digital twins, enabling secure data sharing across supplier and customer networks for capacity, material and constraint visibility.Agent-first, adaptive — Multiagent AI enables planning to shift from “plan and execute” to “sense and adapt”, incorporating physics- based constraints to improve simulations’ realism. — Collaborative multi-agent AI links machines, lines, workers and partner sites through digital twins and physics- grounded models to coordinate operational signals in real time. — AI agents and planners collaboratively adjust production slots, routes and schedules in near real time, learning from disruptions across connected sites.Living, autonomous networks — Autonomous, self-healing planning networks operate across multi-plant and multi-enterprise ecosystems. — Physical AI continuously blends real world feedback (e.g. machine vibrations, material behaviour, energy usage, workforce ergonomics) with accelerated simulation to train autonomous decision agents (e.g. MEGA Blueprint).4 This enables planning networks that self-heal, rebalance flows and optimize under real constraints, not just digital ones. — These networks continuously synchronize with ecosystem- wide demand and supply shifts. Objectives NOW (0-2 years) NEAR (3-5 years) NEXT (5+ years)PLANNING THEME 1 Intelligent Industrial Operations Outlook 2026 14