Right-Sizing, Smart-Routing: Automation Finds Its Footing in the Warehouse and the Last Mile

In the fast-moving, margin-thin world of e-commerce, automation is no longer a luxury—it's an operational imperative. From the warehouse floor to the curbside drop, companies are rethinking legacy systems to meet rising consumer expectations. But amid the automation gold rush, the question remains: how much of this is transformation, and how much is tech theater?

Two recent conversations—one with a UK-based tool retailer, the other with a VP at a last-mile AI platform—offer a nuanced picture of what real automation looks like in 2025, and what it doesn’t (yet) promise.

At UK Planet Tools, a fast-growing online power-tool retailer, the story starts in the packing room. “This machine was exactly what we wanted,” says Managing Director Bohdan Hrystayenko of their investment in Sparck Technologies’ CVP Impack auto-boxing system. Capable of producing 500 right-sized packages an hour, the system has trimmed cardboard use by 30% and reduced reliance on manual labor. “Now we don’t begin [packing] until 8am,” he adds. “Thanks to the CVP Impack, just two operators can manage 250–300 packages per hour.”

It’s a textbook example of proactive tech adoption—one that’s already paying off in lowered consumables, fewer staff hours, and higher throughput. But even Bohdan admits they were punching below the machine’s weight at the time of installation. The early investment was a bet on growth—and not every operation has the luxury or confidence to leap ahead of capacity.

Image credit: Sparck Technologies

On the other end of the logistics pipeline, Megan Murphy, VP of Product at Circuit for Teams, says the adoption curve for AI in last-mile delivery is still uneven. “Couriers need to benefit from efficiency gains when making the leap from analog to digital before enough trust is earned to let AI make key business decisions,” she notes. While tools like real-time route optimization and predictive failure analysis are becoming more sophisticated, Murphy points out that inconsistent data quality across the supply chain remains a major bottleneck.

“There’s still no solved handoff,” she admits, pointing to the fragile connection between customer checkout forms and the labyrinth of first-, mid-, and last-mile systems. For all the power of AI, a wrong postal code still means a missed delivery.

Interestingly, both stories—one about box size, the other about delivery timing—point to the same operational truth: automation only works when it addresses real-world friction. Whether that’s a packing bottleneck on a Monday morning or an incorrect address on a Friday night, the effectiveness of AI and robotics is limited less by ambition than by execution.

Still, there are glimmers of what a more connected, intelligent fulfillment chain could look like. At UK Planet Tools, Hrystayenko is already imagining higher peaks. “Had we installed the CVP before the autumn peak,” he reflects, “we could have done 1.5x the business.”

For now, automation in logistics sits at a crossroads between promise and pragmatism. The tools are here. The use cases are emerging. But adoption—like everything in supply chain—is happening in increments, not leaps. The hype may be heady, but as these companies show, the real gains come from sizing technology not to the pitch deck, but to the problem.