Washrooms in airports, transport hubs, campuses, stadiums and large offices are cleaned to a timetable that has no relationship to how heavily they were used. Occupancy, consumable levels, bin fill and air quality reported in real time replace that timetable with demand — and the measurable result is usually fewer cleaning hours and a higher satisfaction score, because the hours land where they are needed.
Every sensor added is a battery, a mounting point and a maintenance obligation. The set below is what earns its place; anything beyond it needs a reason.
| Signal | How | What it changes |
|---|---|---|
| Footfall and occupancy | Infrared people counter at entry, or door counter | Cleaning triggered on a use threshold rather than a clock |
| Cubicle occupancy | Door position or passive infrared per cubicle | Live availability display; queue and dwell analytics |
| Consumable level | Level sensing in soap, paper and sanitiser dispensers | Refill before empty, not after a complaint |
| Bin fill | Ultrasonic or time-of-flight in the waste bin | Removes the most common visible failure in a public washroom |
| Air quality | Ammonia, hydrogen sulphide, VOC, humidity and temperature | Objective evidence for a complaint that is otherwise subjective |
| Leak and flood | Water presence sensor at floor level and under fittings | Minutes rather than hours to detect an overflow |
| User feedback | Physical panel at the exit, four buttons, no app | A satisfaction signal tied to a time and a location |
| Water and flush volume | Pulse metering on the supply | Consumption baseline; a rising floor indicates a running valve |
The feedback panel. Occupancy tells you how many people were in the room; the panel tells you what they thought of it. Correlating the two is what turns a cleaning schedule into an argument you can take to a facilities budget meeting.
The chain is short. It fails at the last link far more often than at the first — a reading that does not become a task is a reading that changed nothing.
Coin cell or primary lithium, no mains at the cubicle. Reporting is event-driven with a heartbeat rather than a fixed interval, because interval is what determines whether the device lasts a contract term or a season.
LoRaWAN or BLE depending on building size and what already exists. Tiled walls, steel partitions, plumbing risers and water make washrooms genuinely difficult radio environments, so placement is surveyed rather than assumed.
Multi-protocol, buffering through uplink outages so an outage costs latency rather than data. One gateway typically covers a floor or a wing depending on construction.
Ingest, device registry, per-room thresholds and escalation. Thresholds are per room, because a concourse washroom and a back-office one behave nothing alike and a single global rule fits neither.
The output is a job in the cleaning team’s existing system with a location, a reason and a time, plus a live display where one is wanted. An email into a shared inbox is where these deployments go to die.
Response time against threshold, consumable stock-outs, complaint correlation and cleaning hours per thousand visits — the numbers a facilities contract is actually renewed or renegotiated on.
Sensing a bin level is a solved problem. These five are what determine whether a deployment is still running in year three.
There is no mains power behind a cubicle door, and a site visit to change a cell costs more than the sensor. Reporting policy — event-driven with a heartbeat, not fixed interval — is what decides whether the fleet survives the contract term. It is a design decision made at architecture, not a setting tuned afterwards.
Ceramic, water, metal partitions and plumbing risers make a washroom one of the worst small-space RF environments in a building. Coverage is surveyed on site before a gateway count is quoted, because a desk estimate here is reliably wrong.
Every device gets sprayed with cleaning chemical as a matter of routine. Enclosure sealing, material compatibility and mounting all have to assume chemical attack and high humidity, and vandal resistance matters in any public venue.
Ammonia and hydrogen sulphide sensors age and need recalibration; low-cost VOC sensors report that something changed, not what. Stating that honestly at specification — and planning the calibration interval — prevents an air quality dashboard that is quietly wrong by month eight.
Thresholds set too tight generate more tasks than the team can service, and within a fortnight everyone ignores the system. Thresholds are tuned per room against observed use, with a deliberately quiet first month, because a system nobody trusts is worse than no system.
Faststream builds computer vision systems, and vision would answer several of the questions on this page more precisely than the sensors that are used instead. It is not offered here, it will not be quoted, and it will not be built if it is asked for.
Everything on this page is non-imaging. Infrared beam counting registers that a body crossed a line. Door sensors register position. Passive infrared registers motion in a zone. None of them produces an image, none can be made to produce an image, and no device in the room stores or transmits one.
Occupancy data is aggregated per room and per interval, not per person. Nothing here identifies anyone, tracks anyone between rooms, or can be reconstructed to do so.
These systems are bought on a business case and lost on operations. Proving both on a single floor costs a fraction of a site-wide rollout that has to be defended afterwards.
| Stage | What happens | Indicative duration |
|---|---|---|
| Site survey | Radio propagation, mounting points, power, existing systems and cleaning workflow observed as it actually runs | 1–2 weeks |
| Pilot floor | One floor or wing instrumented, thresholds deliberately loose, no alerting yet | 4–6 weeks |
| Threshold tuning | Observed use compared against feedback scores; thresholds set per room | 2–4 weeks |
| Workflow integration | Alerts routed into the cleaning team’s existing system as tasks | 2–4 weeks |
| Rollout | Remaining floors, provisioning at scale, fleet management and staged updates | By site size |
| Operation | Battery monitoring, calibration schedule, threshold review and reporting | Continuous |
Faststream engineers the sensing, gateway, connectivity, platform and integration layers, and designs custom hardware where an off-the-shelf sensor does not fit the fixture. Dispensers, fixtures and the cleaning service itself come from your existing suppliers — we integrate with them rather than replace them.
A set of battery-powered sensors reporting occupancy and footfall, consumable levels, bin fill, air quality, leaks and user feedback over a low-power radio network, so cleaning and restocking are driven by actual use rather than by a fixed timetable.
No. Faststream does not use cameras in washrooms under any circumstances. All sensing is non-imaging — infrared beam counting, door position and passive infrared motion — and no device in the room captures, stores or transmits an image. Occupancy is aggregated per room and interval, never per person.
It depends far more on reporting policy than on cell chemistry. Event-driven reporting with a periodic heartbeat, rather than a fixed short interval, is what allows a fleet to last a contract term. Because there is no mains power at a cubicle, and a site visit costs more than the sensor, this is settled at architecture.
Usually LoRaWAN or BLE, chosen against building size, construction and what infrastructure already exists. Washrooms are difficult RF environments — tile, water, steel partitions and plumbing risers — so gateway placement is established by site survey rather than estimated.
It should not. The output is a task in the cleaning team's existing system with a location, a reason and a time. An alert that generates an email into a shared inbox is the most common reason these deployments quietly stop being used.
Yes, and that is the normal case. Faststream engineers the sensing, connectivity, platform and integration layers and designs custom hardware where an off-the-shelf sensor does not fit a fixture. Dispensers, fixtures and cleaning services stay with your existing suppliers.
Airports and transport hubs, stadiums and venues, shopping centres, university and hospital campuses, and large office estates — anywhere footfall is high, uneven, and poorly matched to a fixed cleaning timetable.
Sensors, gateways, links and device cloud.
SOLUTIONBuildings, utilities and estates.
PRODUCTThe gateway layer, standard or custom.
KNOWLEDGEWhy LoRaWAN or BLE here.
PRODUCTProvisioning and fleet operation.
INDUSTRYSector context.
Tell us the specification, the constraint and the deadline. Programmes that cross silicon, radio, embedded and AI are where Faststream is strongest.