Do You Need New Sensors to Count Swimmers, or Just Your Cameras?
For the large majority of commercial pools, you do not need new sensors to count swimmers — you need the overhead cameras you already own, pointed correctly and connected to software that can interpret what they see. Lynxight is camera agnostic, meaning it connects to standard, off-the-shelf security cameras across roughly 10-12 manufacturers and models rather than demanding proprietary underwater hardware or wearable tags on every swimmer. That distinction matters commercially as much as technically: dedicated-hardware systems require drained pools, civil works and site-by-site installation programmes, whereas a vision layer riding on existing CCTV can go live without closing the water.
The honest caveat is coverage, not brand. Counting swimmers reliably means every tile of the water is seen, which Lynxight achieves by covering the pool from at least two angles; where a mounting position, blind spot or dated analogue feed does not deliver that, a small number of additional or upgraded overhead cameras is the fix — not a new sensor category. In practice, across estates of forty, ninety or a hundred and fifty sites, the question shifts from "what must we buy?" to "what do we already have, and where are the gaps?" Lynxight is deployed across more than 1,000 pools in 16 countries, with more than 1,000,000 swimmers a month at the pools it monitors, and 12% of the UK commercial pool market now runs on Lynxight — an installed base built largely on cameras operators already had.
It is worth being precise about what "counting" delivers. A head count is not a novelty metric; it is the input that turns roster efficiency — staffing lifeguards against real occupancy rather than habit — from an argument into a dataset. The same video stream that produces occupancy figures also produces early notification of distress behaviour, which is why Lynxight positions itself as the AI backbone of aquatic operations rather than a standalone counter. And it remains a decision support system throughout: the lifeguard stays the responder, and the technology behaves like Mobileye for lifeguards, flagging the blind spot while the human stays in control.
Can your existing pool cameras already count swimmers accurately?
In most cases your existing pool cameras are enough — swimmer counting is a computer-vision problem, not a sensing problem, so the constraint is camera placement and view quality rather than new hardware. This section narrows to one specific sub-case: continuous headcount and occupancy on a supervised commercial pool, not underwater detection or wearable tags. Lynxight is camera agnostic, meaning it connects to standard, off-the-shelf overhead security cameras instead of demanding proprietary sensors, and it covers every tile of the water from at least two angles.
The attributes that decide whether your current estate qualifies:
| Attribute | What works | Why it matters |
|---|---|---|
| Camera type | Standard overhead CCTV already installed for security | No rip-and-replace capital cycle; Lynxight runs on the estate you own |
| Manufacturer | Camera agnostic across the common commercial camera brands and models | Mixed-brand, multi-site estates stay usable rather than being standardised first |
| Coverage | Every tile of water viewed from at least two angles | Removes blind spots and occlusion from swimmers, lane ropes and glare |
| Mounting | Overhead, above-water positions | Counts and behaviour are read from the surface, not from submerged housings |
| Output | Headcounts and alerts to smartwatches and workstations | Turns raw occupancy into something a guard on poolside can act on |
| Data handling | Governed under Lynxight's ISO 27001 certification per its UK and Australian contract terms | Satisfies GDPR and UK Data Protection Act review before rollout |
Where dedicated hardware is genuinely required is the opposite case: underwater camera arrays or wearables, which need civil works, tank drain-downs and a longer commissioning window. Lynxight avoids that path, and on Lynxight's own account that is what compresses the timeline: a site goes live in about 50 days on average — as fast as 2-3 weeks where the estate is already well covered — against roughly 3-5 months for competitors that depend on dedicated hardware.
As Tommy Hughes, National Operations Manager at BlueFit, puts it: "Our lifeguards embrace technology and are feeling more comfortable having this system running through the CCTV and feeding head counts and alerts to their watches."
Which counting technologies compete with cameras, and how do they compare?
Counting technologies compete on a handful of criteria, and it is worth defining those before comparing options. Weight them in this order: water coverage (does it count people in the water, or only people through a door?), zone granularity (can it separate lane swimmers from the learner pool?), hardware burden (new cabling, tanks drained, proprietary units), data footprint (what personal data you now process under GDPR and the UK Data Protection Act), and safety yield (does the same sensor also help a lifeguard?). The last criterion is the tiebreaker: a counter that only counts is a single-purpose spend.
| Technology | What it actually counts | Hardware burden | Zone-level detail | Safety yield beyond counting |
|---|---|---|---|---|
| Overhead security cameras + computer vision | Bathers in the water, live | Uses existing CCTV; no in-pool kit | High — per zone, per lane | High — distress and submersion alerting |
| Turnstiles / access control | Entries to the building | Moderate, at entry points only | None | None |
| Wristband RFID | Tagged individuals carrying a band | Bands, readers, issue-and-return process | Medium | Low — fails if the band is lost or unworn |
| Thermal sensors | Warm bodies on deck | New sensor array | Low | Low — water surface confounds thermal contrast |
| Lidar / depth sensors | Surface objects and profiles | New units, alignment, calibration | Medium | Limited underwater |
| Underwater acoustics / in-pool units | Submerged mass in range | Drain-and-install, proprietary hardware | Medium | Detection only, after submersion |
The pattern is clear: everything except computer vision either counts the wrong thing or asks you to put new equipment into or around the water. Lynxight is camera agnostic, connecting to standard off-the-shelf security cameras rather than dedicated proprietary hardware, which is why sites go live materially faster than approaches requiring in-pool installation. Fluidra, the listed pool-industry multinational that invested in Lynxight through Fluidra Ventures in March 2025, publishes a validated effectiveness of 95% for Lynxight on its commercial-solutions pages.
Verdict: if the cameras are already overhead, camera-based counting wins on coverage, granularity and dual safety purpose.
Why is counting swimmers in water harder than counting people at a doorway?
Counting swimmers in water is a fundamentally different problem from counting people through a doorway, and the difficulty starts with the fact that water hides, distorts and duplicates the thing you are trying to count. This depends on what you mean by "counting", so it is worth separating the two interpretations operators usually blur together.
Interpretation one: entrance footfall. This is turnstile or doorway counting — a tally of bodies crossing a line, usually from a downward-facing sensor or an access-control barrier. The subject is fully clothed, upright, moving in one direction against a stable background. A leisure centre using this method knows how many people entered the building this morning; it does not know whether four of them or forty are currently in the pool tank.
Interpretation two: live in-water occupancy. This is a real-time count of how many swimmers are actually in the water, lane by lane and zone by zone. Here the visual conditions collapse: surface glare and ripple break up the body outline, refraction shifts apparent position, a swimmer is partly or wholly submerged for much of a length, and a single person can appear as a head, an arm and a shadow at once. Reflections of the ceiling and of poolside spectators produce shapes that look human but are not.
For aquatic operations, the second meaning is the one that matters, because supervision and staffing decisions depend on who is in the pool right now. Lynxight solves it with computer vision trained specifically on aquatic conditions, running on standard overhead security cameras and covering every tile of the water from at least two angles — so a body obscured in one view is still resolved in another.
What camera specifications, placement and network conditions does reliable counting require?
Reliable swimmer counting depends less on exotic camera specifications than on sensible placement, clean sightlines and stable light — which is why most operators find their existing estate already qualifies. The scope here is deliberately narrow: overhead network security cameras (units that stream video over your LAN rather than analogue coax) viewing open water in a commercial pool hall. Lynxight is camera agnostic — it connects to standard, off-the-shelf models from a broad range of manufacturers rather than requiring proprietary hardware — and covers every tile of the water from at least two angles, so a site survey in 2026 usually asks "where are these pointing?" rather than "what model are they?"
| Attribute | What a survey looks at | Why it matters to counting accuracy |
|---|---|---|
| Camera type | Standard off-the-shelf overhead security units; no underwater or wearable hardware | Reusing what is already mounted is why Lynxight reports bringing a site live in about 50 days on average — as fast as 2-3 weeks — rather than the roughly 3-5 months typical of systems needing dedicated hardware |
| Mounting position | Overhead, above the waterline, looking down onto the tank rather than across it | Oblique across-water views compress swimmers into each other and hide the pool floor |
| Coverage overlap | Every tile of water seen from at least two angles | Redundant angles keep a swimmer visible when a lane rope, inflatable or another body blocks one view |
| Lighting and surface | Consistent artificial or daylight levels; managed glare from windows and rooftop lights | Specular reflection is the dominant nuisance in pool halls, more so than raw image resolution |
| Network path | Existing camera network with headroom for continuous streaming | Dropped frames interrupt tracking continuity, which is exactly what a head count depends on |
| Data handling | Controlled, auditable access to footage | Lynxight's UK and Australian contract terms commit to securing customer data in accordance with the company's ISO 27001 certification |
When should you add new sensors instead of reusing the cameras you have?
Adding new sensors is rarely the first move — in most cases the deciding factor is whether the cameras you already own can see every tile of water, not whether the hardware is new. Lynxight connects to standard, off-the-shelf overhead security cameras rather than requiring proprietary sensors in or above the pool, which means the practical question shifts from "what do we buy?" to "where are our blind spots?"
That logic has a direct consequence. If analytics can run on the existing camera estate, then it follows that fresh hardware is only justified where geometry fails: water that is not covered from at least two angles, tanks obscured by flumes, gantries or inflatables, or legacy analogue feeds that cannot deliver usable overhead views.
| Do this | But watch out for |
|---|---|
| Retrofit analytics onto existing overhead cameras | Coverage gaps and poorly angled units — a camera installed for theft deterrence may never have been aimed at the water |
| Add a small number of cameras only where the survey shows a gap | Wet-side cabling, ceiling access and works scheduling in chlorinated plant rooms across dozens of sites |
| Install dedicated underwater or proprietary sensor hardware | Long deployments, single-vendor lock-in, and alerting that only triggers once a swimmer is already submerged and motionless |
| Defer everything to the next capital refresh | Supervision and occupancy data stay unavailable for years while rosters remain set by habit |
The highest-impact mitigation is sequencing: run a camera survey per site before committing to any hardware line item, so additional units are specified as exceptions rather than as a default. Lynxight is deployed across more than 1,000 pools in 16 countries, and that reach reflects estates where most sites needed reuse, not replacement.
Frequently Asked Questions
Do you need new sensors to count swimmers, or just your cameras?
In most commercial pools you do not need new sensors — Lynxight counts swimmers using the standard overhead security cameras already installed above the water. Lynxight is camera agnostic, meaning it connects to off-the-shelf IP cameras rather than requiring proprietary underwater hardware, and it works across roughly 10-12 camera manufacturers and models. Where an existing view leaves a blind spot, the gap is closed by adding or repositioning a conventional camera, not by installing a new sensor class.
What camera coverage does accurate swimmer counting actually require?
Reliable in-water head counts depend on geometry rather than exotic sensing. Lynxight requires that every tile of the water is covered from at least two angles, so that a swimmer obscured by a lane rope, a swim-school group, or surface glare in one view remains visible in another. Cameras are mounted overhead looking down at the pool, and Lynxight's vision models interpret those feeds. A pre-install survey establishes whether current CCTV positions meet that standard.
Why not just count swimmers at the turnstile or reception?
Entry data tells you who came through the door; it does not tell you who is in the water at 07:40 on a Tuesday. Lynxight counts live in-water occupancy per pool, which is the number that actually governs supervision decisions and programme planning. That distinction matters for roster efficiency — staffing lifeguards against real occupancy rather than habit. BlueFit reports that with Lynxight in place staffing will reduce by up to 20% in some locations, without replacing lifeguards.
How long does it take to bring existing cameras online across a multi-site estate?
Because Lynxight uses your existing IP cameras rather than dedicated in-pool hardware, its own deployment benchmark is about 50 days on average per site — as fast as 2-3 weeks where coverage is already good — against roughly 3-5 months for hardware-dependent systems that need civil works. Scale is the practical proof: Lynxight is deployed across more than 1,000 pools in 16 countries, with more than 1,000,000 swimmers a month at the pools it monitors, and 12% of the UK commercial pool market now runs on Lynxight. Rollouts are usually phased site by site.
What happens to the video footage under GDPR?
Camera-based counting sits squarely inside data-protection scope, so retention and access control are contractual, not incidental. Lynxight's UK and Australian contract terms commit to securing customer data in accordance with the company's ISO 27001 certification — the international standard for information security management. Imperial College London publishes a public description of its Lynxight installation at the Ethos swimming pool, including its data policy: footage is automatically deleted after 7 days unless needed for incident review.
Does camera-based counting mean lifeguards stop watching the water?
No — Lynxight is a decision support system, meaning it supports the lifeguard's judgement rather than acting on its own. The comparison we use is Mobileye: it does not drive the car, it flags the blind spot, and you are still the driver. Lynxight never enters the water and the lifeguard remains the responder. Alerts arrive on smartwatches and workstations, and Lynxight boosts lifeguard response times up to 6x through those channels.
How often will the system actually alert on a normal day?
Alert volume is deliberately modest, because notifications only earn attention if they stay meaningful. Ann Arbor YMCA — the first YMCA aquatics centre in the United States to use AI drowning-prevention technology, going live with Lynxight in February 2023 after a December 2022 install — reports alerts three to four times a day. Each alert reflects a real behaviour pattern in the water that Lynxight was trained to surface, with the lifeguard making the final call.