The brief
The operator was replacing a legacy warehouse WLAN. Handheld scanners and truck terminals still needed 2.4 GHz; office users and cameras wanted 5 GHz. Existing APs sat on the outer walls. Mid-aisle RSSI collapsed once palleted stock went in, and scan guns reassociated constantly at the far ends.
A predictive survey from the CAD would have been cheaper, but racking attenuation is the thing drawings lie about. We scoped an AP on a Stick survey plus a Sidekick 2 walk of neighbouring SSIDs.
What we measured
We raised a representative AP to the proposed mount height in three aisle types: empty, half stocked and fully plated. At each height we captured RSSI, SNR and retry rates on 2.4 and 5 GHz, then walked the same path without the stick AP to see the interferers already in the shed (forklift radios, neighbouring units, leftover SSIDs).
The useful finding was not “need more APs”. It was that 5 GHz died two bays sooner than the predictive model, while 2.4 GHz looked fine on a heatmap and still dropped scans because of CCI from the perimeter APs all sharing a handful of channels.
What we recommended
- Move from wall-mount to aisle-aligned runs over the racking, with APs looking down the lane rather than across steel.
- Separate scanner SSIDs onto 2.4 GHz with a tight channel plan; keep office and camera SSIDs on 5 GHz.
- Lower TX power on 2.4 GHz so cells stop overlapping three aisles away.
- Bill of materials with AP count, mount height, antenna type and switch PoE budget.
Result
The installer cabled to the APoS points. A later validation walk showed secondary coverage at the aisle ends and scan-gun roam events that no longer dumped sessions. Same building, fewer APs than a “one every 20 metres on the wall” guess.
If you have a similar shed in Southampton, Portsmouth or anywhere we survey warehouses, send the floor plan and racking height. We will tell you whether you need APoS or a predictive first.