MHI Solutions Improving Supply Chain Performance
How AI Helps Supply Chains Predict and Mitigate Disruptions
THE WORLD OF supply chains looks radically different from what it was a few years ago, according to Aadil Kazmi, head of artificial intelligence (AI) at MHI member Infios.
He noted that while software systems built over the past three decades remain resilient, data‑driven optimizers capable of scaling to millions of events per minute, they were never designed to be real‑time, actionable or truly intelligent. “At the core of every scenario is, if something happens, do ‘this, that’ in that order,” Kazmi explained, referring to the rigid, rules‑based logic that underpins traditional systems.
The traditional model of scenario planning requires organizations to build and maintain a workflow for every possible situation—a costly and time‑consuming process to execute, manage and optimize using today’s technology. AI, he said, changes that equation by unlocking the ability to scale the number of scenarios covered both reliably and efficiently.
To achieve that level of agility, organizations must connect their systems of record—their data—to agentic action platforms such as agents and workflow automation tools, and then deploy those systems where the work gets done.
Despite the promise, AI adoption in the supply chain has been slower than in other industries. Kazmi said this is because AI adoption often begins on the operational side of the supply chain before filtering down to shippers and manufacturers. “As the supply side of the supply chain—carriers, brokers, technology providers—adopt AI to reduce costs or improve operational excellence, the demand side—shippers and others—look to replicate these efficiency gains,” he said.
As a counterpoint, Bill Denbigh, vice president, GTM strategy at MHI member Tecsys, said his company is seeing the opposite trend: rapid AI adoption across supply chain operations.
Addressing the role AI can play in helping companies maintain a competitive edge against supply chain market volatility and change, Denbigh points out that the current state of AI adoption in logistics—such as in warehouse and transportation—is around resource optimization and adaptability, with AI predicting approaches and actions that make more efficient use of resources.
Lila Fridley, AI vice president and general manager of MHI member C3, said scenario planning and improving supply chain resilience are two of the highest‑impact AI use cases. “Scenario planning historically was accomplished with rules‑based systems, which are rigid, unresponsive and hard to set up—requiring manually writing if‑then and optimization logic,” she said.
AI scenario planning, she explained, can generate scenario models itself, such as producing the code to perform a scenario analysis, alleviating the manual effort of coding heuristics based on user knowledge.
Additionally, the scope of the scenario types that can be modeled is more flexible and dynamic, adapting to changing real‑world conditions, Fridley said. “Users can run scenarios for new, previously unencountered situations, like dramatic changes to supply, demand, capacity or externalities such as changes to regulatory environments and geopolitics,” she said, adding that AI‑driven scenario planning enables supply chain leaders to improve resilience for short‑ and long‑term decisions.
Fridley also noted that AI systems can automatically run hundreds or thousands of scenarios in parallel, applying reasoning and logic to compare and synthesize results. That capability accelerates decision‑making and enhances the expertise applied to evaluating complex, multi‑horizon trade‑offs—such as balancing short‑term cost savings from lower‑priced outsourced materials with long‑term supply stability, and understanding the ripple effects of supplier delays lasting one, two or 10 weeks.
Fridley said that supply chain market volatility introduces short‑term disruptions—leading to stockouts and missed demand—as well as long‑term operational structural shifts and fundamental changes in operating costs and strategy. With AI, both challenges can be addressed.
“When responding to rapid, unpredictable disruptions like labor shortages, energy price swings, weather events and consumer behavior change, AI applications help supply chain managers identify optimal plans and courses of action to mitigate the interruption,” Fridley said.
When faced with stockouts in high‑volume or strategically important regions, firms are using AI models to determine the best way to reallocate inventory. Others use AI models to forecast and predict energy load, informing production plans and schedules to help maintain a consistent supply and proactively avoid peak-priced days, she said.