Warehouse Wi-Fi and Mobile Devices in Freezers

The freezer was surveyed in March, before the season's stock arrived, and every aisle showed a healthy signal. By October the racks are full of frozen product, reach-truck drivers lose their session halfway down aisle 14, and pickers coming out to the dock wipe fog off their scanners with a glove before each scan. The WMS shows the same pick confirmed twice, and the site blames the network vendor.

Each of those failures has a cause that can be calculated in advance: radio loss through loaded racks, a handover that has about a second to finish at the door, and condensation that follows from the dock's dew point. Plan the three together, with the arithmetic written down, and a freezer can run on mobile computers as reliably as the ambient floor. This guide sets out that arithmetic and the decisions that follow from it.

• What this covers: where to put access points in and around an insulated cold room, path-loss arithmetic, roaming across the freezer door, condensation and dew point, device heaters and batteries, spectrum rules in the US and India, and acceptance tests.
• What it adds: a path-loss and dew-point calculation block, a dated fog-and-duplicate incident, and a dollar and rupee comparison of heated devices against staging procedures.

Why freezer aisles defeat a coverage plan drawn for the ambient floor

A freezer is a metal box filled with water in solid form, and both properties work against radio. Treat it as a separate radio space with its own access points and its own survey, rather than as an extension of the ambient floor plan.

Cold rooms are commonly built from insulated sandwich panels with steel facings. Sheet metal reflects radio energy, so an access point on the ambient side of the wall contributes little once the doors close. Inside, pallets of frozen product absorb and scatter the signal, and the loss changes with the season as racks fill and empty. That is why a survey taken in an empty or half-empty room overstates coverage.

• Panel walls and ceiling: plan coverage from inside the room; count any signal leaking through doors as a bonus, not a design input.
• Loaded racks: survey at peak inventory or simulate it with product-filled pallets in the test aisles.
• High-bay geometry: reach and very narrow aisle trucks put the device 8 to 12 m up, so measure at height as well as at hand level.
• Doors and strip curtains: the door is where most sessions break, because the device leaves one radio space and enters another within a second or two.

The nearest page on this site, our voice picking and RF scanner comparison, compares how pickers interact with devices. This article sits underneath that choice: whichever device you pick, it needs coverage, a working handover and protection from the cold.

Access-point placement: inside the room, at the wall or through it

Put access points inside the freezer unless a survey proves otherwise. The real decision is how to keep them within their operating temperature: buy units rated for the room, house standard units in heated enclosures, or keep the radio outside and pass only antennas through.

Start from the access point datasheet's operating temperature range. If the room runs below it, the unit needs either a cold-rated design or an enclosure with a thermostatically controlled heater. Each enclosure needs a cable entry through the panel that the refrigeration contractor seals, and its heater adds to the power budget, which may exceed what the network switch supplies over the data cable.

• Cold-rated units inside: simplest to maintain; check the rated minimum against the coldest setpoint, including defrost overshoot.
• Standard units in heated enclosures: widest choice of models; adds heater power, a thermostat to monitor and a door seal to keep.
• Radio outside, antennas inside: keeps electronics warm; adds cable loss and panel penetrations, and is not available for 6 GHz low-power indoor units in the US, which must have integrated antennas.
• Mounting height: above the racking, pointed down the aisle, or at aisle ends; avoid positions behind evaporator coils and their fans.

For the US, FCC order 20-51 opened the 5.925 to 7.125 GHz band to low-power indoor access points in April 2020, but its definition of an indoor access point requires a wired power supply, an integrated antenna, no battery and no weatherized enclosure. Ask the maker whether a heated housing is acceptable before choosing 6 GHz for a freezer. In India, 5 GHz wireless access systems are licence-exempt under G.S.R. 1048(E) of 2018; confirm the Department of Telecommunications rules in force before buying 6 GHz-only clients.

Path-loss arithmetic for a loaded freezer aisle

Do the arithmetic before the survey so that you know what the survey should find. Free-space loss sets the floor; loaded racks, panel reflections and body loss add to it, and only measurement with product in place tells you how much.

Recommendation ITU-R P.525 gives free-space basic transmission loss between isotropic antennas as 32.4 plus 20 log of frequency in megahertz plus 20 log of distance in kilometres. At 5,500 MHz that is about 67 dB at 10 m, 75 dB at 25 m and 79 dB at 40 m. At 2,437 MHz the same distances give about 60, 68 and 72 dB, which is why 2.4 GHz reaches further but carries less and is more crowded.

Freezer link budget and condensation check (illustrative inputs)

Free-space loss, ITU-R P.525 eq. (4):
  L_fs = 32.4 + 20*log10(f_MHz) + 20*log10(d_km)            [dB]
  f = 5500 MHz, d = 0.025 km  ->  L_fs = 75.2 dB

Received signal at the device:
  RSSI = EIRP_ap - L_fs - L_racks - L_body + G_client          [dBm]
  EIRP_ap  = 17 dBm   (access point setting, illustrative)
  L_racks  = 2 faces * 10 dB = 20 dB   (measure on site)
  L_body   = 3 dB     (operator between device and AP)
  G_client = 0 dBi
  RSSI = 17 - 75.2 - 20 - 3 + 0 = -81.2 dBm   vs design target -67 dBm

Dew point of dock air, WMO-No. 8 Magnus coefficients (water):
  gamma = ln(RH/100) + 17.62*T/(243.12 + T)
  T_dew = 243.12*gamma / (17.62 - gamma)                       [deg C]
  T = 32, RH = 75  ->  T_dew = 27.0 C
  T = 24, RH = 50  ->  T_dew = 12.9 C

Condensation risk if device surface temperature < T_dew (frost if < 0 C).

The rack loss of 10 dB per frozen pallet face and the 17 dBm transmit setting are illustrative; the point is the shape of the answer. A device two loaded faces away from the access point at 25 m misses a -67 dBm target by about 14 dB, so the plan needs an access point closer to that aisle, or one at each aisle end, rather than more power from the existing one.

Freezer aisle coverage for warehouse mobile computers should be budgeted from ITU-R P.525 equation (4): free-space loss at 5,500 MHz over 25 m is about 75 dB, and at an illustrative 10 dB per frozen pallet face, two faces leave a device about 14 dB short of a -67 dBm design target.

Roaming across the freezer door at reach-truck speed

The door is where sessions fail. A truck at 12 km/h moves about 3.3 m a second, so it crosses a 4 m door zone in roughly 1.2 seconds, and the device must complete its handover to the next access point inside that window.

Three IEEE 802.11 amendments, now folded into the base standard, address this. 802.11r, Fast Basic Service Set Transition, shortens the security exchange at each handover. 802.11k gives the device a list of neighbouring access points so it does not have to scan every channel, and 802.11v lets the network suggest a better access point. Enable them only after testing every device model, because older clients may refuse to join a network that advertises a feature they cannot use.

• Overlap at each door: place an access point on each side of every freezer door so that both are heard in the door zone, not just one on each side of a wall.
• Consistent settings: the same network name, security type and band on both sides, so the device does not reconnect from scratch.
• Roaming trigger: set the device to start looking for a neighbour before the signal collapses, using the maker's roaming threshold setting.
• Application tolerance: the WMS client should hold a scan locally for a few seconds and resend it with the same transaction identifier, so a retry never creates a second confirmation.

The last point belongs to the application team, not the network team. A handover that takes 300 milliseconds is invisible to a picker, but a WMS session that times out and replays the last confirmation produces exactly the double pick in the opening example. Our page on offline scan continuity covers longer outages; the same identifier rule protects short ones.

Roaming at a freezer door for warehouse mobile computers is a 1.2-second problem at 12 km/h, so IEEE 802.11r fast transition, overlapping access points on both sides of the door and WMS scan resends that reuse one transaction identifier together prevent dropped sessions and duplicate pick confirmations.

Condensation and frost: dew-point arithmetic at the dock door

Condensation is not a device defect; it is physics. Any surface colder than the dew point of the surrounding air collects water, and a device straight out of a -25 °C room is far colder than any dock's dew point.

• Conditioned US dock: using the Magnus coefficients in the WMO guide, air at 24 °C and 50% relative humidity has a dew point of about 13 °C, so a -25 °C device sits 38 °C below it.
• Humid Indian dock: air at 32 °C and 75%, common in the monsoon season, has a dew point of about 27 °C, leaving the same device 52 °C below it and fogging far more heavily and for longer.
• Return to the freezer: a device taken back into the cold while still wet freezes the film onto the scan window.

1. 19:42, illustrative evening shift: picker P-27 leaves freezer aisle 14 after 40 minutes with mobile computer MC-0815 and starts staging order SO-554210 at dock door 6, where the air is 31 °C and 72% humidity.
2. 19:43: the screen and scan window fog within seconds. Two scans of the outer carton fail; the picker wipes the window with a glove and scans again, and the WMS records pick confirmation PC-998120.
3. 19:44: the device's session to the dock access point is still re-establishing after the door crossing. The client resends the confirmation with a new transaction number, and the WMS posts PC-998121 for the same line.
4. 19:58: the picker returns to the freezer. The condensed film freezes, scanning slows across the next 12 picks, and the line count shows a short shipment against the doubled confirmation.
5. Controlled version: the device goes onto a staging shelf at the dock edge for its warm-up, the picker uses a dock-side spare, and the WMS rejects the second confirmation because it carries the original transaction identifier.

Heaters, batteries and screens on freezer-rated mobile computers

A freezer device is a different specification, not a standard one in a sleeve. Specify the operating temperature, the heated parts, the input method for gloved hands and the battery rules before comparing prices.

• Rated temperature: the device's operating range should cover the coldest room plus the time a user actually spends in it, with margin for defrost cycles.
• Heated scan window and display: reduces fog and frost on the surfaces that matter most; check the heater's effect on battery runtime.
• Glove-friendly input: physical keys or a touchscreen mode tuned for gloves, since insulated gloves rarely work with standard touch settings.
• Battery capacity in the cold: runtime drops in cold conditions, so plan battery swaps per shift from a test in the actual room, not the datasheet's room-temperature figure.
• Charging only when warm: research using neutron diffraction on working cells found that low-temperature charging plates metallic lithium on the anode and that only part of it reverses, so charge only within the pack's rated charge temperature.
• Charging room: chargers sit in an ambient room near the freezer, and batteries rest there before charging, never in the dock air lock.

Battery charging for freezer mobile computers belongs in an ambient charging room, because research by Zinth and colleagues in the Journal of Power Sources (2014) found that low-temperature charging plates metallic lithium on the anode and only part of that plating reverses, permanently reducing capacity.

Heated devices or a staging routine: costing it in dollars and rupees

Whether heated freezer devices pay for themselves depends heavily on labour cost. The same fog problem that justifies a premium device in a US DC within months may be cheaper to solve with staging shelves and spare devices in an Indian one.

The illustrative case is a 6,000 m² freezer at -25 °C served by six access points and 40 mobile computers, with 25 pickers making 6 freezer exits a shift over 2 shifts, and each fog event costing about 4 minutes of slowed scanning. That is 20 labour hours a day. Prices, wages and counts are assumptions; replace them with your own quotes and payroll.

No device removes all fog delay, so both payback figures are best cases. The contrast is the point: in the Indian case, a staging routine with spares costs about the same as one year of fog delay and is far cheaper than the device premium, while in the US case the premium device wins quickly. The access-point positions are needed in both.

Checks the network and WMS enforce, and where tools only advise

Some freezer controls can be set once and enforced by the wireless controller or the WMS; others need a person to read a survey or a temperature log. Keep that split explicit so nobody assumes a dashboard is making decisions it is not.

The same split applies to sensors. Our guide to warehouse temperature mapping covers placing loggers in the room; if those loggers use the Wi-Fi network, include them in the survey and the access point plan, since they sit in the same racks and suffer the same loss.

Survey, rollout and acceptance checklist for a freezer network

Accept a freezer network only after it passes tests under the conditions that break it: full racks, closed doors, moving trucks and humid docks. The checklist below sets out what to test and what counts as a pass.

If you are planning a freezer extension or a device refresh and want the survey, the WMS resend rules and the staging routine designed together, outline your site in a request for proposal. More warehouse design topics are in the logistics and supply chain library.