Headlines love humanoid robots, but the real wins in supply chain are happening on the warehouse floor, pallet by pallet, move by move. We sit down with Dr. Matt Waller, Brian Nachtigall from ArcBest Vaux , and Brad Umphres from Deloitte to unpack how autonomous forklifts with human-in-the-loop teleoperation cut costs, boost safety, and create the clean data that WMS and AI engines need to make faster, better decisions.
We start with the practical: why pairing sensors and remote teleoperation with proven forklift platforms beats ripping and replacing, and how that approach thrives in messy, changing DC environments. From there, we map the ripple effects; accurate, time-stamped pallet moves reduce lost inventory, unlock smarter slotting, and fuel interleaving and pick-path optimization. That visibility compounds across the network as TMS models incorporate forecasted weather and traffic, while drones accelerate cycle counts. The result is speed where it counts: speed of decisions. When your data is trustworthy, you move goods faster without accelerating bad processes.
The conversation then zooms out to the big picture. We connect rising productivity, resilient labor, and moderating inflation with capital deepening in AI and automation, a shift already visible in earnings and capex trends. For leaders deciding how to act, the playbook is clear: choose high-impact use cases, set hard metrics, educate teams, and commit beyond pilots. We share patterns that work, top-down sponsorship, real change management, and a “burn the ships” mindset that turns tools into adoption and pilots into programs.
If you’re weighing where autonomous forklifts fit in your DC, how to get clean inventory data without a rebuild, or what it takes to scale AI from a test bay to network-wide impact, this conversation is your blueprint. Subscribe, share with your ops and IT leaders, and leave a review with the one use case you’re ready to implement next.
More About this Episode
AI & Supply Chain: The War for Speed
AI in supply chain has moved from “interesting” to “urgent.” The reason is speed: speed of decisions, speed of execution, and speed to the customer. Companies that learn how to apply AI to real operations; warehouses, transportation, inventory, and planning, are starting to pull away from competitors who are still stuck evaluating.
The shift from hype to usable value
A couple years ago, AI in supply chain was mostly pilots, demos, and big promises. What’s changing now is that the tech is landing on specific, measurable use cases. With LLMs improving and “agentic” AI becoming more practical, organizations can finally put AI to work in ways that impact cost and service, not just dashboards.
The pattern is clear: what was hyped two years ago is becoming reality now, and what’s hyped today is likely to show up in production soon, if companies take action instead of waiting for certainty.
Warehouses are the front line
The most immediate, tangible gains are showing up in warehouses because warehouses are where speed and accuracy collide. If a product is in the wrong place, or you don’t trust your data, everything downstream slows down: picking, loading, routing, store replenishment, last-mile delivery.
Automation plus AI helps in a very practical way:
- It improves visibility: where pallets actually are, not where someone last scanned them.
- It improves integrity: systems get better data because movement is recorded reliably.
- It improves safety: fewer unpredictable movements in congested environments.
- It reduces cost: less time searching, fewer errors, fewer wasted touches.
One of the strongest themes is that “better data makes faster decisions possible.” When you trust the information, you don’t hesitate, you execute.
Why autonomous forklifts matter right now
Forget humanoid robots for a second. The near-term win is autonomous and human-in-the-loop forklifts moving pallets. The forklift is already a proven tool. The innovation is layering sensors, machine vision, and software on top so pallet movement becomes trackable, repeatable, and increasingly automated.
A human-in-the-loop approach matters because real warehouses change constantly. Layouts shift. Processes evolve. Exceptions happen. The model that works is: automate what’s repeatable, keep humans available for what’s messy, and use that operational learning to automate more over time.
It’s a faster path to value than trying to redesign an entire facility around perfect automation.
Speed isn’t just physical, it’s cognitive
A lot of supply chain conversations focus on physical speed: faster picking, faster loading, faster delivery. AI adds a second layer: faster decision-making.
AI systems can ingest huge amounts of data, weather forecasts, projected traffic, demand shifts, capacity constraints, and turn them into operational choices faster than humans can. That’s how you enable things like same-day delivery promises without chaos: rapid decisions upstream make the downstream experience feel smooth.
When decision latency drops, performance improves.
End-to-end use cases are getting real
Warehouse is the headline, but the momentum is end-to-end:
Demand planning is expanding beyond historical sales to incorporate signals like sentiment and trends. Inventory positioning models are getting smarter about balancing cost and service levels. Transportation optimization is moving from reactive routing to predictive routing using forecasted constraints. Autonomy in trucking is still early, but pilots and advanced driver-assist capabilities are growing.
The key difference now is that these aren’t just “AI ideas.” They’re use cases with owners, metrics, pilots, and rollout plans.
Successful pilots don’t scale by accident
The biggest difference between a pilot that “works” and one that becomes operational is commitment.
When leadership treats a pilot as optional, people try it when they can, and results are mixed. When leadership says “this is the direction,” adoption improves, teams engage, and performance jumps, because the organization adjusts around the tool instead of treating it like a side project.
Scaling AI requires change management: education, clear goals, guardrails, and top-down support. The tech matters, but the human system around it matters just as much.
Don’t automate broken processes
One warning shows up repeatedly: speeding up a bad process only creates faster failure. Before you automate, you have to understand what you’re automating. AI and robotics amplify whatever you already are, good or bad.
That’s why the best implementations start with a clear use case, clear measurement, and a willingness to refine the process as the tech rolls out.
The competitive risk of waiting
This moment looks a lot like the early web era: some companies moved fast and built durable advantages; others waited and lost ground they never got back.
AI and automation are becoming a supply chain baseline. Companies that over-focus on governance and risk avoidance, to the point of paralysis, are likely to fall behind organizations that learn by doing.
The right move isn’t reckless adoption. It's a structured action: start, measure, iterate, expand.
Jobs will change, not vanish
AI will change job descriptions across the supply chain. The work shifts toward oversight, exception handling, system tuning, and continuous improvement. Humans still matter because exceptions, tradeoffs, and judgment calls are everywhere in real operations.
Companies that reskill alongside automation will gain speed without losing stability.
The takeaway: speed is the new edge
The war for speed is not just “faster delivery.” It’s faster learning, faster decisions, and faster execution, powered by better data and smarter tools.
If you’re leading in this space, the message is straightforward: don’t ignore AI, don’t wait for perfect certainty, and don’t assume competitors are standing still. The advantage is going to the companies who lean in now and operationalize AI where it matters most.
