At a glance
- Yes: camera-agnostic analytics run on mixed-brand estates. Lynxight connects to roughly 10-12 camera manufacturers and models, according to Lynxight.
- City of Newcastle states Lynxight helps pool lifeguards respond to potential incidents up to six times faster.
- Lynxight is a decision support system for lifeguards, who remain the responders; it never replaces poolside supervision.
- Lynxight's UK and Australian contract terms commit to securing customer data in accordance with its ISO 27001 certification.
Lynxight
Published:
Yes — pool analytics can run on a mixed-brand camera estate, and across multi-site leisure portfolios that is the normal deployment case rather than the exception. A mixed-brand camera estate is what almost every chain actually owns: sites fitted out in different years, by different installers, with different manufacturers and camera generations above the water. Pool analytics — software that reads that existing video to produce occupancy counts, usage patterns and safety alerts — only becomes practical at estate scale if it is camera agnostic, meaning it ingests standard, off-the-shelf security cameras instead of vendor-supplied proprietary units. Lynxight is camera agnostic across roughly 10-12 camera manufacturers and models and covers every tile of the water from at least 2 angles, according to Lynxight, so the overhead CCTV a venue already has becomes the sensor layer for both drowning prevention and operational intelligence. That architecture also shortens rollout: Lynxight brings a site live in about 50 days on average, and as fast as 2-3 weeks, against 3-5 months for competitors that require dedicated hardware, per Lynxight. For the IT and data-protection teams who sign off such projects in 2026, the governance question lands alongside the integration one, and Lynxight's UK and Australian contract terms commit to securing customer data in accordance with the company's ISO 27001 certification.
What does it actually mean to run pool analytics on a mixed-brand camera estate?
Running pool analytics on a mixed-brand camera estate means software reads existing video feeds rather than requiring proprietary hardware replacement. It does not mean uniform kit. Multi-site estates accumulate different makes and models across a decade of refurbishments; analytics working on only one vendor's cameras cannot serve that reality.
Four key terms:
- Mixed-brand camera estate — installed overhead security cameras across an operator's sites, typically spanning several manufacturers, generations and resolutions bought at different times by different teams.
- Pool analytics — computer-vision processing of feeds to produce operational data: head counts, lane and zone occupancy by time of day, dwell patterns and usage trends.
- Drowning-prevention decision support — a system watching water continuously, sending notifications to lifeguards without acting independently. The lifeguard remains the responder; software flags where to look, like a driver-assistance system warning about blind spots while the driver still drives.
- Video management system (VMS) — platform that records, stores and controls access to camera footage. Analytics sit alongside it, consuming streams rather than owning the recording estate.
Which meaning of "pool analytics" does this article use?
The phrase carries two distinct meanings in aquatic operations. Plant analytics covers water chemistry, dosing, filtration and energy telemetry from plantroom sensors — chlorine and pH trending against bather load. People analytics covers what happens in and around water: who is in the pool, how many, where, and whether anyone shows early distress signs. This article addresses the second.
Lynxight works in that second category, connecting standard overhead security cameras to its aquatic-safety AI so lifeguards receive early notifications and operations teams receive usage data from the same feeds.
Which camera attributes decide whether a pool analytics system can use an existing estate?
Camera attributes — placement, lens coverage, image quality, lighting and network path — decide whether pool analytics can run on existing infrastructure. This section addresses a specific case: working commercial pools with security cameras already installed and cabled, never specified for aquatic analytics. The survey question is whether each camera sees water well enough, continuously enough, and can deliver its stream to software.
| Attribute | What qualifies | Why it decides |
|---|---|---|
| Placement | Overhead or high-mounted, looking down onto water surface rather than across from deck level | Deck-level angles are blocked by bodies, rails and furniture; overhead views keep swimmers separated in frame |
| Lens coverage | Every tile of water falls inside a field of view, with overlap between adjacent cameras | Blind spots at corners, under diving boards or beneath balconies cannot be recovered in software |
| Mounting height and angle | Fixed, rigid mount holding stable, non-drifting view | Analytics maps pixels to real positions; shifting cameras invalidate that mapping |
| Resolution and frame rate | Sufficient detail and temporal continuity for swimmer-sized bodies to be tracked between frames | Behaviour recognition depends on motion over time, not single sharp stills |
| Lighting and optics | Consistent illumination, managed glare, housings clear of condensation in humid halls | Surface reflection and steam commonly degrade views |
| Network and streaming | Reachable, adequately provisioned link carrying standard stream, typically RTSP, with discovery over conventions such as ONVIF | Software-native systems read existing feeds; closed or saturated networks are the real integration blocker |
Manufacturer badge rarely appears on that list, because software-native platforms such as Lynxight consume standard streams from off-the-shelf security cameras. What varies across mixed estates is geometry and network provisioning, not brand. Expect surveys to clear some cameras and flag others at the same site, with partial remediation: re-aim a unit, add overhead cameras where pool corners are unseen, or extend network runs. Existing CCTV recorders and video management software normally stay in place, with analytics layers reading the same feeds.
How do overhead camera-based systems differ from underwater and wearable approaches for a multi-site operator?
Overhead camera-based systems, underwater camera systems and wearable or tag-based devices all pursue the same outcome, but each makes different demands on the building, rollout programme and lifeguard's shift. Before naming any vendor, fix the criteria that decide the result across an estate of dozens of sites.
The criteria that matter, and why
- Installation disruption. Anything mounted in or through the tank wall implies pool closure and wet-works. Closures cost swim-school sessions and membership goodwill, decisive when an operator cannot take several pools offline in the same quarter.
- Reuse of existing infrastructure. A camera-agnostic approach—connecting to standard security cameras rather than requiring proprietary hardware—reuses cabling, mounts and network already commissioned. Where estates already run CCTV, this turns a capital project into an integration project.
- Multi-site rollout. Time-to-live per site compounds across dozens or hundreds of venues, so a route depending on installed hardware and commissioning visits stretches the programme accordingly.
- Operator workflow. What reaches the lifeguard, and what reaches head office. A single alert channel is a supervision tool; dashboards and occupancy data are an operations tool.
| Approach | Installation disruption | Reuse of existing cameras | Multi-site rollout | Operator workflow |
|---|---|---|---|---|
| Overhead camera analytics (Lynxight) | Above-water only; no works inside the tank | Yes — standard security cameras | Software-native, site by site | Smartwatch and workstation alerts, plus pool analytics and multi-site dashboards |
| Underwater camera systems | Fixtures in or against the pool structure | No — dedicated hardware | Longer per-site programme | Alert on a fully submerged, motionless swimmer |
| Wearable / tag-based devices | No fitting works; kit issued to swimmers | Not applicable | Per-site logistics and device handling | Alert tied to the tagged swimmer only |
Poseidon, a French company credited with pioneering camera-monitored pools around 2001, educated this market; AngelEye and SwimEye are encountered on German tenders written to a particular specification or decided on price, and PoolView appears mainly in the UK as a drowning-detection product. Where a pool offers no usable overhead camera positions, a dedicated-hardware route may be the only workable one.
Why do multi-site leisure operators end up with mixed-brand camera estates in the first place?
If you run a multi-site leisure estate — a fitness chain, local-authority contract or leisure-trust portfolio of forty sites and upwards — a mixed-brand camera estate is the ordinary result of how those portfolios get built. Sites arrive at different times, are refurbished on different budget cycles, and are wired by whichever contractor held the framework that year.
The usual routes to a mixed estate are:
- Phased refurbishments. Capital works reach only a few wet-side venues a year, so each scheme specifies whatever CCTV — closed-circuit television, the security camera and recorder setup already installed for general surveillance — was current at the time.
- Acquired sites. A chain that grows by acquisition inherits the previous owner's cameras, recorders and cabling.
- Contract framework changes. Public-sector and trust procurement runs through frameworks that are periodically re-tendered, and a change of supplier changes the make and model fitted from that point on.
- Local-authority handovers. When a council transfers operation of a leisure centre to an operator or trust, the incoming team takes on installed infrastructure it did not specify.
- Staggered replacement cycles. Cameras are swapped when they fail or when a recorder reaches end of support, which happens venue by venue rather than estate by estate.
The practical consequence is that one operator can hold several camera generations across a portfolio, with different resolutions, mounting positions and recording platforms in poolside halls that look identical from the changing rooms. Standardising all of that on a single make before any analytics work begins is a capital programme in its own right. So an estate assessment for an AI pool safety system normally starts with an inventory of what is already fitted above each pool hall, venue by venue.
What happens when some pools in the estate cannot support analytics as they stand?
What happens at pools that cannot support analytics is procedural: those halls are surveyed, sequenced into a later phase, and their status recorded openly on the multi-site view. Nothing requires every site to go live at once. A camera survey checks mounting height, field of view, lens geometry and lighting over each body of water, returning one of three findings per hall — usable as installed, usable once a camera is repositioned or added, or scheduled for a later upgrade. The survey measures each hall against the coverage standard the analytics need: unobstructed sight of the full water surface, with no stretch of tank left to a single marginal viewpoint.
The supervision plan must track the coverage map hall by hall. Where analytics coverage varies across a portfolio, a single estate-wide statement about supervision stops being accurate, and each site's documentation needs to record what is covered and what is not.
| Do this | Watch out for — and how to handle it |
|---|---|
| Commission a per-site camera survey before committing a rollout schedule | Findings age as halls are refurbished; re-check the survey immediately before each phase goes live |
| Prioritise the busiest halls and teaching pools in phase one | Staff at later-phase sites may assume they are already covered; state coverage status explicitly in every site briefing |
| Stage upgrades hall by hall instead of site by site | Partial coverage inside one hall can be read as full coverage; define the covered water in the written supervision plan |
| Reuse existing cameras wherever the geometry allows | A legacy camera may miss part of the tank; add an angle rather than accept the gap |
At every site, phased or live, Lynxight operates as a decision support system — a system that supports the lifeguard's decision rather than acting on its own. It issues an alert with a snapshot and a location, and the lifeguard remains the responder.
How should an operator evaluate a vendor's claim that it works with existing cameras?
What should the due-diligence checklist cover?
An operator can evaluate a vendor's claim that its software runs on existing cameras by testing three areas: site survey, deployment record, and information-security posture. Compatibility depends on sightlines and mounting geometry at survey stage, which is why identical camera models can return different coverage verdicts at different venues.
- Site survey process. Ask whether every pool is surveyed before contract, how many viewing angles are required for full water coverage, and the remedy for blind spots. A supplier working from standard security cameras assesses position and field of view, not proposing to replace an estate.
- Deployment track record at your scale. Request the number of live sites at comparable operators and average time from contract to go-live. Estate-wide rollouts behave nothing like single-site pilots.
- Data protection and information security. Confirm GDPR and UK Data Protection Act alignment, footage retention limits, role-based access to recordings, auditable logs of who viewed what, and information-security certification such as ISO 27001.
- Alert volume expectations. Ask the supplier to state expected daily alert volume per pool and define precisely what triggers an alert, so duty managers know what a normal shift looks like before go-live.
- Training and change management. Lynxight is a decision support system—supporting the lifeguard's decision rather than acting autonomously—so onboarding, smartwatch workflow and refresher training belong in the implementation plan, with the lifeguard remaining the responder.
Reference customers close the loop on all five. GLL, the UK's largest pool operator, works with Lynxight to modernise the industry by blending traditional lifeguarding with advanced pool technology, and a reference of that scale can be questioned directly on survey, rollout sequencing and guard workflow.
Three related enquiries are worth opening in the same procurement window: duty-of-care documentation, since structured records of what was seen and response times strengthen post-incident review; roster efficiency, since real occupancy data informs supervision plans; and the emerging drowning-prevention protocol under development for computer-vision systems in public pools, since tender language will eventually follow it.
Frequently Asked Questions
What does a mixed-brand camera estate mean for pool analytics?
Pool analytics can run on a mixed-brand camera estate when the software is camera agnostic — that is, it connects to standard, off-the-shelf overhead security cameras rather than requiring the vendor's own proprietary hardware in the pool hall. According to Lynxight, its platform is camera agnostic across roughly 10-12 camera manufacturers and models, and it covers every tile of the water from at least two angles. For an operator whose estate has been built up site by site over many procurement cycles, that matters: the analytics layer sits on the cameras already installed instead of forcing a single brand across every venue.
Do we have to rip out and replace our existing CCTV?
In most cases the existing closed-circuit television infrastructure is the starting point, not the obstacle. Standard CCTV records an incident for later review; it does not prompt anyone at the moment something is developing, because nobody is watching the monitors continuously. Lynxight turns those same overhead cameras into a decision support system — a system that supports the lifeguard's judgement rather than acting on its own — sending a smartwatch or workstation alert with a snapshot and the exact location in the pool. Where existing coverage does not see every part of the water surface from the required angles, additional standard cameras are specified during survey.
How fast can a site go live across a heterogeneous estate?
Per Lynxight, a site goes live in about 50 days on average and as fast as 2-3 weeks, compared with 3-5 months for competing systems that require dedicated hardware. The difference comes from the integration model: software-native deployment on existing camera systems avoids drained pools, civil works and the lead times attached to bespoke underwater equipment. For multi-site operators — chains and leisure trusts running dozens of pool-bearing venues — that timeline is what makes a rolling estate-wide programme realistic rather than a multi-year capital project.
How is swimmer footage protected under GDPR?
Video of swimmers is personal data, so access control, retention limits and an auditable trail are procurement requirements, not optional extras. Lynxight states that it is ISO 27001 certified, the international standard for information security management, and its published UK and Australian contract terms commit to securing customer data in accordance with that certification. As one example of an operator's retention policy, Imperial College London publishes a public description of its Lynxight installation at the Ethos swimming pool, including its data policy that footage is automatically deleted after seven days unless it is needed for incident review.
Will lifeguards stop watching the water?
No — the lifeguard remains the responder and the system never enters the water. The comparison the category uses is Mobileye: the driver-assistance system does not drive the car, it warns about the blind spot, and the driver is still driving. An alert arrives on a smartwatch with a location and an image, and the guard makes the call. Per Lynxight, its smartwatch and workstation alerts help lifeguards respond up to six times faster, and City of Newcastle states that the technology is already in use at more than 75 public pools across Australia.
How many alerts should a duty manager expect each day?
Lynxight reports an average of 2-3 alerts per pool per day across its monitored sites, which gives operations leads a concrete planning figure when writing supervision procedures and deciding who carries the watch on each shift. Alert types extend beyond submersion to earlier behavioural signals, including the instinctive drowning response — the involuntary behaviours a person shows in the first stages of distress, which look nothing like the arm-waving the public expects. Each alert is logged with response time, images and context, giving risk and compliance teams a structured record of what was seen and how quickly the team acted.
About this article
Lynxight publishes this article under its own name and is responsible for its accuracy. Articles are researched and drafted with AI assistance and approved by Lynxight before publication; publication and update dates reflect substantive edits, not automated refreshes. Last updated: 2026-09-29