Private 5G replaces a tangled Wi-Fi mesh at a 900k sq ft DC

Context

A national 3PL was preparing to deploy autonomous mobile robots (AMRs) across a 900,000 sq ft distribution center. The existing Wi-Fi mesh — installed in three different generations over a decade — was failing AMR handoff requirements during pilot testing. Latency spikes during roaming were causing AMRs to pause mid-aisle, costing throughput and operator confidence.

The problem

The 3PL had received two proposals: one for a wholesale Wi-Fi refresh ($2.2M) and one for a Private 5G overlay ($3.8M). Internal opinion was split. Engineering favored Private 5G; operations was nervous about training and supportability; finance wanted the cheaper option. No one had measurement data that supported any of the three positions.

Our approach

• We pulled AMR telemetry from the pilot, normalized handoff latencies by location, and overlaid them on a heat map. The roaming failures clustered in two specific zones — both areas where the Wi-Fi mesh had been extended in 2019 without re-tuning the channel plan.
• We modeled three remediation paths: a targeted Wi-Fi fix, a full Wi-Fi refresh, and a Private 5G overlay. We sized the BOM and total cost of ownership for each, including the operations team's training and support burden.
• Targeted Wi-Fi work would have addressed the existing AMR pilot, but couldn't accommodate the planned doubling of the AMR fleet within 18 months. Private 5G could, with margin.
• We wrote a vendor-neutral RFP for the Private 5G build, including operational SLAs and AMR handoff acceptance criteria. Three vendors bid; we normalized responses and benchmarked against published list.
• We helped the 3PL select an integrator (not the highest-bid, not the lowest) and stayed in the room through cutover and acceptance testing.

The outcome

AMR handoff latency dropped 63% vs. the legacy Wi-Fi mesh. The 3PL doubled its AMR fleet within the planned window without revisiting the network. The Wi-Fi mesh was retained for general staff devices; Private 5G carried AMR traffic on a separate slice. Total installed cost came in 14% under the original vendor's Private 5G quote.

"We had three vendor proposals and three different stories. They were the only ones who showed up with measurement data instead of an opinion. By the time we picked a path, we knew exactly why."

Measurable results