Which band should barcode scanners run on: 2.4, 5, or 6 GHz?+
For mid-market warehouses in 2026 the answer is almost always 5 GHz as the primary with 6 GHz for new Wi-Fi 6E or Wi-Fi 7 handhelds, keeping 2.4 GHz only for legacy label printers and IoT. 2.4 GHz is too saturated and too prone to microwave and motor interference to carry picking traffic. The right split depends on your actual device roster: a fleet of older Zebra MC9300s may still need a 2.4 GHz plan, while a new Zebra TC53 or Honeywell CT45 fleet runs happily on 5 and 6 GHz.
Do DFS channels help or hurt in a warehouse?+
They help capacity but hurt reliability if used naively. DFS channels (52 to 144 in most regions) unlock more non-overlapping 5 GHz capacity, but a radar hit forces an immediate channel change that can drop a voice-pick session. The right design uses DFS channels on fixed-location APs away from airports and weather radar, and keeps non-DFS channels on the APs covering voice-pick and VoIP zones. Every DFS decision in the survey is explicit, not left to the controller is auto algorithm.
How does racking density change AP placement?+
Dramatically. Nine-metre racking filled with metal, cardboard, and liquid attenuates 5 GHz by 10 to 25 dB per rack, so an AP mounted above the racks sprays most of its energy sideways and almost nothing reaches the aisle floor. The right pattern in dense racking is aisle-centre mounting with downtilt or patch antennas pointed along the aisle, not a uniform ceiling grid. We model attenuation from your CAD and validate it on site, aisle by aisle.
What about cold storage and freezer rooms?+
Cold-storage condensation kills consumer-grade APs in weeks and most mid-range enterprise APs within a year. The right AP for a freezer room has an operating range down to minus 30 Celsius, an IP-rated enclosure, and often a cable gland rather than a sealed-port variant. We also budget for condensation drip loops and placement that avoids the thermal shock of a defrost cycle. If the current AP in your freezer is the same model as the one in your office, it will fail.
What about PoE budget on our existing switches?+
A Wi-Fi 6E or Wi-Fi 7 AP typically draws 25 to 35 W under full load, so a 48-port switch with a 740 W PoE budget cannot run forty-eight APs plus phones. The BOM we deliver flags every switch where the PoE budget is tight after the survey and includes the upgrade cost, if any, in the written estimate. We do not ship a design that assumes the switch will magically find more watts.
How do we get smooth roaming for voice-pick between aisles?+
Voice-pick needs sub-50 ms roam time with no audio gap, which only comes from a correctly configured 802.11k, v, and r stack combined with cell overlap in the 15 to 20 percent range at minimum data rate. We tune minimum data rates to force client stickiness breakage, enable 802.11k and v so the AP tells the client where to go next, and 802.11r for fast transition where the devices support it. Vocollect A700 and A730 headsets, Zebra WT6300 wrist terminals, and Honeywell CT45 handhelds all behave differently; the design accounts for the slowest-roaming device on your floor.
Is Wi-Fi 6E or Wi-Fi 7 worth the refresh now?+
For a greenfield warehouse, yes. For a brownfield site, it depends on how much of your handheld fleet actually supports 6 GHz. A survey that recommends Wi-Fi 7 without inventorying your scanner fleet is marketing, not engineering. We inventory every device SKU, map the ones that support 6 GHz, and only recommend the refresh when the fleet justifies it or the next three-year roadmap does.
Ekahau, iBwave, or Hamina: which tool do you use?+
All three, depending on the site. Ekahau is our default for mid-size warehouses because the predictive engine and the active-survey rig are the most mature. iBwave shines on multi-floor manufacturing with heavy mezzanines and in-building DAS alongside WiFi. Hamina is strong when the design has to iterate quickly with multiple vendor options. The deliverable is the same regardless: calibrated model, documented wall materials, and a report your next vendor can read.
Which handhelds and headsets do you test against?+
Zebra TC53, TC58, TC73, MC9400, WT6300, and the legacy MC9300 line; Honeywell CT45, CT47, CT60, and CK65; Datalogic Memor 20 and Skorpio X5; Vocollect A700 and A730 voice-pick headsets. We also test against Cisco 8821 and Spectralink VoIP handsets where in scope. If you run a device we have not touched before, we buy or borrow one for the survey; nothing goes into the report without a real device measurement.
How do we fix the nulls that forklift masts create?+
Metal masts moving through the warehouse create temporary nulls at ten-metre height, exactly where the forklift terminal sits. The fix is a combination of cell density (so the terminal always has a second viable AP), antenna selection (patch antennas pointed down the aisle rather than omni from the ceiling), and in some sites a dedicated 5 GHz SSID for forklifts with a slightly higher minimum data rate so the terminal roams off a dying cell faster. We model this in the predictive phase and validate it on site with the forklift actually moving.
Should scanners live on the same VLAN as everything else?+
No. The design we deliver puts scanners, voice-pick, VoIP, cameras, badge readers, and guest traffic on separate VLANs with specific QoS tagging, and the switch and firewall configuration is part of the deliverable. Mixing scanners and guest WiFi on one VLAN is how a single malware incident takes picking offline, and it is also how DHCP exhaustion crashes a shift. The segmentation is not optional in our designs.
What is actually in the survey report, and how long does the on-site survey take?+
The report contains the predictive heatmaps (RSSI, SNR, data rate, roaming), the measured passive and active heatmaps, interference and rogue inventory, PoE budget assessment, AP placement plan with mounting detail, channel and power plan, VLAN and QoS plan, vendor-neutral BOM, and a validation protocol for post-install sign-off. On-site survey time is typically one day for sites up to 10,000 square metres, two to three days for 10,000 to 40,000 square metres, and four to five days for 40,000 to 100,000 square metres, including cold-storage, mezzanine, and yard where applicable.
