Skip to content

a western Japan fulfillment center serving packaged consumer brands

How a Consumer-Goods Logistics Hub Doubled Picking Productivity With AMRs

A western Japan 3PL hub cut picker walking in half and nearly doubled throughput with a 12-robot AMR fleet after a one-day proof of concept.

~2x
Picking output
50%
Staff need
50%
Daily walking
12
AMR units

Based on a documented real-world deployment. Figures are from public reporting; the organization is not named.

Tall warehouse shelving lines a wide picking aisle with cartons stacked on pallets.
Photo: Daniel Andraski

Manual carts across a wide floor

Pickers at a western Japan fulfillment center serving packaged consumer brands pushed heavy carts through long aisles every shift. Site leaders reported daily walking near 20,000 steps per person, with fatigue showing up fastest on cart pulls.

A major consumer-goods client asked the third-party operator to raise productivity and trim labor without freezing outbound service. Labor pressure had been building for years, so another manual workaround was not an option.

  • Cart-based picking strained shoulders and backs
  • Walking distance dominated each shift
  • Client demand required measurable productivity gains

AMRs after a year of vetting

The operator spent about a year comparing fixed automation. Conveyors and guided vehicles looked costly and rigid for a hub that still reshuffles SKUs. Goods-to-person AMRs won because other company sites already ran them successfully.

Managers and floor staff from the client toured a live reference site, then the hub ran a proof of concept with two to three days of on-site prep and a single test day. The larger XL tote format was validated against real product dimensions and aisle clearances before the July 2024 production start with twelve units.

  • Benchmark multiple AMR vendors
  • Host client walkthrough at a running site
  • Run a one-day proof of concept
  • Train pickers on large-screen interfaces
Warehouse supervisors review plans together beside stacked cartons.
Photo: Felicity Tai

Measured gains on the floor

A warehouse associate walks an open aisle without pulling a loaded cart.
Photo: Tiger Lily

After go-live, leaders estimate daily walking per picker fell by about half from the prior 20,000-step norm. Picking productivity nearly doubled, and the headcount required to run picking dropped by half compared with the old cart workflow.

Floor staff now walk to pick locations without dragging loaded carts across the building. Feedback highlighted easier shifts, especially for team members who struggled with the old cart pulls. Screens and guided next-task prompts shortened the learning curve after a few formal training sessions.

What US 3PLs can borrow

The pattern is familiar on this side of the Pacific: when walking and cart handling eat the shift, goods-to-person AMRs buy back hours without pouring concrete for fixed automation. Service Robot Co. stays vendor-neutral, maps the pick paths, finances buy or lease options, and keeps units running through RoboCare once they are live.

A short proof of concept on your SKUs beats a slide deck. We help you set success metrics, stage the fleet beside your WMS or host system, and train leads before you commit to a full count.

Frequently asked questions

No. They execute horizontal moves while your WMS or host system still owns inventory and order logic. Integration work focuses on pick guidance, hand-offs, and exception paths rather than ripping out existing software.

More case studies

Free site assessment. We tell you what actually works before you spend a dollar.