For supply chain leaders, autonomy starts with a practical question: Can the plan keep learning from what’s actually happening in stores, warehouses, and distribution networks, or does it stall the moment reality moves off-script?
The urgency is growing. Gartner predicts that by 2031, 60% of supply chain disruptions will be resolved without human intervention as AI enables increasingly autonomous operations. Autonomous supply chains need more than AI alone. They need the data, demand intelligence, and planning capabilities to see change earlier, weigh the options, and act within defined guardrails.
“There’s always been this desire to do things automatically,” said Nicholas Wegman, Ph.D., senior director artificial intelligence scientist at Zebra Technologies. “But that wasn’t really possible with older technologies. It was very rules-based or programmatic, and it couldn’t adapt.”
The difference now is adaptability. Earlier automation worked best when conditions stayed within the rules. Once an unusual scenario appeared, systems often pushed the problem back to people. AI gives supply chain teams a more flexible layer of decision support, one that can help manage routine decisions while elevating the exceptions that need experienced judgment.
Autonomy Is Only as Strong as What It Can See
The system can only act on what it can access and interpret. AI can analyze more information than a person can process manually, but it still needs the right inputs.
“Just because you have all this great AI technology, that does not mean that you can ignore having the right foundational data,” Wegman said. “Those AI systems are going to look at that data to make their decisions.”
Those inputs include the structured data companies already rely on, such as ERP, shipment, and inventory data. But AI readiness also requires a broader view of information: promotions, emails, commercial conversations, notes, and other context that may have lived in people’s heads or inboxes.
In the past, companies may have left that information outside the planning process because older models couldn’t easily use it. AI changes the value of those overlooked inputs and gives organizations a reason to bring more of that context into view.
Planning Still Starts With Demand
Across retail and consumer packaged goods, demand remains the starting point for inventory, replenishment, and service decisions. Consumer behavior is shifting quickly, shaped by price pressure, changing channels, short-lived trends, and new expectations.
“What is driving that supply chain is consumer demand,” Wegman said. “If you don’t know what the customers want, you’re not going to be able to fulfill it.”
That makes demand forecasting and analysis central to autonomy. Better demand intelligence helps teams understand what people are buying, why they’re buying, where they’re buying, and how those patterns are changing. It gives AI agents a stronger basis for recommendations and helps planning systems respond with more precision as conditions move.
War Rooms Signal a Planning Gap
When disruption overwhelms the planning process, teams often have to move outside their normal systems to solve the problem manually.
“If a problem happens and the only way to satisfy that problem or to solve it is to stop using your existing tools, pull everything into a spreadsheet, get in a war room and go make a bunch of manual adjustments, then your supply chain planning process is not being proactive enough,” Wegman said.
A more autonomous planning process keeps more of that work inside the system. Some disruptions will always arrive unexpectedly. The question is whether the system has enough intelligence, data, and planning logic to recommend or take the next best action.
If a shipment is late, demand spikes, or inventory falls short, the system needs a reliable view of demand and inventory to decide what should happen next. Stronger forecasting and planning capabilities make that response faster, more consistent, and less dependent on manual workarounds.
Human Judgment Moves to the Exceptions
As routine decisions become more automated, human judgment becomes more important in the right places.
Wegman described AI as a force multiplier: another set of eyes on decisions teams may not have time to review one by one. In practical terms, that means letting agents handle lower-risk decisions within clear guardrails around order size, decision size, or other defined thresholds.
That keeps people in the process while changing where their effort goes. Instead of acting as manual troubleshooters for every exception, planners can focus on the harder cases, guide the AI, and improve how it handles similar issues over time.
The result is a more strategic role for supply chain leaders. They set the guardrails, review the decisions that matter most, and help shape the system’s judgment as it learns.
Execution Data Closes the Loop
For Zebra, autonomy depends on connecting planning to what’s happening on the ground. Demand forecasting and inventory optimization create the base. Execution feedback makes it stronger.
“If you just stick a commercially available model on top of a poor process, you’re still going to get poor results,” Wegman said.
Zebra’s advantage, he added, comes from capturing execution data through technologies such as barcode scanning, RFID, and other frontline data sources, then feeding that reality back into future plans. That closes a gap that has historically made planning feel disconnected from what happened downstream.
“It’s one thing to plan,” Wegman said. “It’s another thing to plan, execute, and then have that execution help inform your future planning.”
That loop is where autonomous operations begin to take shape. Scan events, RFID reads, and other frontline signals can flow back into forecasts and inventory models, giving planning teams a clearer view of what happened, where conditions changed, and what the system should account for next.
A Smarter Path to Autonomous Planning
Zebra’s Demand Forecasting & Analysis and Supply Chain Planning capabilities help organizations connect demand intelligence, planning logic, and execution feedback. As supply chains move toward greater autonomy, that connection can help teams detect disruption earlier, reduce manual intervention, and make better decisions as conditions change.