There is no single best smart roof leak detection system, because two different technologies are sold under the same name. Conductive or impedance layers on or just under the membrane can find the breach itself, to about a square metre, but have to be built into a new roof and need a conductive path beneath the membrane. Wireless moisture sensors tell you water has arrived somewhere in the build-up, can usually be retrofitted, and are cheaper to deploy across a large area. Two things matter more than the choice between them. First, any roof where the waterproofing will be covered benefits from permanent leak detection, because nobody can inspect a covered membrane; which roofs that means is in should you specify leak detection in this roof?. Second, on a covered roof you open up where it is wet whichever system reported it, so the specification is the zone size an alarm must resolve to, not the technology. Choose the zone size first, then the technology, then the product.
Disclosure: Vector has a commercial arrangement with Sensor Innovation: where we design a sensor array, the design is registered with Sensor Innovation and the client can buy the hardware at a discount. We do not sell or distribute the equipment. That is a potential bias, and we state it so you can weigh it. We have assessed Sensor Innovation’s into® system on the same criteria, from the same kind of public evidence, as every other system here, and its limitations are stated as plainly. Ben Hickman is a co-author of CIRIA C817 and the PCA’s buried roofs guidance, both cited below.
This page is drawn from our October 2026 review of the market: twenty-odd providers, their datasheets and case studies, the academic literature, and the UK standards. Where a claim comes from the manufacturer, we say so.
The short answer, by roof
| Your situation | Technology that fits | Systems to look at |
|---|---|---|
| Any new roof where the waterproofing will be covered (green, blue, podium, buried, inverted under ballast or paving) | Permanent leak detection, zoned so that opening one zone is an affordable repair. Point sensors fit most build-ups; a membrane-level grid locates the breach but must be built in and needs a conductive path | Sensor Innovation into®, Tector, Sika Roof Monitoring System, VILPE Sense; grids: ProGeo smartex mx, Detec PermaScan-I, SikaRoof Control (none with a UK installed base) |
| Exposed warm roof over a sensitive space (data hall, archive, hospital), or a large area where early warning is wanted | Wireless point sensors in the build-up (finds the wet zone) | Sensor Innovation into®, Tector, Sika Roof Monitoring System, VILPE Sense |
| Existing roof, leak history, can’t strip it | Retrofit point sensors | Tector, VILPE, SIHGA Monitorix, FLOscout, Sensor Innovation |
| CLT or mass timber deck | Timber moisture probes, plus moisture sensors in the build-up | Tector Woody, Sensor Innovation Sensor Wood, Senzomatic, OmniSense |
| FM Global-insured site | A system with FM 7745 approval | Detec PermaScan-C |
| Warm blue roof | Permanent leak detection should be considered (CIRIA C817 §3.2.3); the waterproofing is covered, so specify it | Point sensors on the vapour control layer; a membrane-level grid where the build-up allows and the client wants the breach located |
Two technologies, two different questions
A conductive layer at membrane level answers “where is the hole?”. A conductive contact layer or tape sits directly under or on the waterproofing. When water bridges a breach, it completes an electrical path, and the system reports where. ProGeo claims about 1 m² for smartex mx; Detec’s PermaScan-I sits on the membrane under overburden. Grids laid lower down answer a different question. Detec’s PermaScan-C, SMT FutureCast (conventional) and ProGeo smartex dm sit on the vapour control layer, under the insulation, so they report which cell has taken water, not where the membrane failed. PermaScan-C was tested on a 1.2 m × 1.2 m grid, about 1.5 m² per cell.
A wireless point sensor answers “has water arrived here?”. A battery sensor sits in the insulation or on the vapour control layer, measures humidity, temperature and in most cases liquid water, and reports by radio. It cannot see the breach. Water can travel some distance through a build-up before it reaches a sensor, so the alarm tells you which zone is wet, not where the membrane failed.
The physics behind electrical methods has hard limits worth knowing before you specify anything. They need water above the membrane and a conductive path below it. Black EPDM, metal-faced felt, decks electrically isolated by a vapour control layer, and timber decks without an added conductive layer all defeat or degrade them (CIRIA C538 §5.3; ASTM D7877). NHBC Chapter 7.1 accepts visual inspection with seam probing for EPDM and foil-faced bitumen for the same reason. A conductive underlay is sometimes offered to make such a build-up testable. Our view is that it adds little: on an inverted roof the membrane already sits on concrete, and on any covered roof the wet zone is where you open up, whichever layer reported it. If a build-up cannot be tested, change the build-up or agree a different QA method; do not rely on an underlay.
Moisture sensors have their own hard problem: telling a leak from normal moisture. SINTEF’s research puts it plainly — normal seasonal redistribution of moisture in roof insulation can produce very high humidity readings at the top or bottom of the insulation. CIRIA C538 noted in 2000 that leak-detection systems are prone to false alarms from condensation. The main point-sensor vendors — Sensor Innovation, Tector, VILPE and Sika — each describe software, weather-data comparison or modelling to cope with this. None has published a detection or false-alarm rate.
The systems, compared
| System | What it senses | Where it sits | Locates to | Retrofit | Connectivity and battery (vendor claims) | Independent evidence | UK presence |
|---|---|---|---|---|---|---|---|
| Sensor Innovation into® (Norway) | Humidity, temperature, liquid water (resistance); timber moisture (Sensor Wood) | In the insulation between vapour barrier and membrane; at drains; in timber | The sensor zone | Listed by vendor; method not published | ~200 m radio to gateway; gateway on LTE-M; up to 30 years (Plus, Wood), 10 years (standard) | Peer-reviewed SINTEF field study (2024) of its sensors, warning method developed with the vendor; a BREEAM approved-innovation entry (2023) covers an advanced leak detection concept and does not name the vendor | Specified by Vector on UK projects (see disclosure); no independent UK distributor found |
| Tector (Denmark, formerly Woodsense) | “Nylon moisture content”, humidity, temperature; timber moisture (Woody) | On the vapour control layer, under the insulation; 1 per 20–25 m² | Approximate, from which sensors respond | Yes | LoRaWAN to a SIM gateway; 25 years, conditional on subscription | Co-authored research with Teknologisk Institut on moisture behaviour; no detection-rate data | UK customers incl. Jesus College Oxford (Northgate House) |
| VILPE Sense (Finland) | Humidity and temperature; mould index | In the insulation, ~4–5 m grid | The sensor zone; slow (days) | Needs the roof opened at each sensor | Indoor base station on mobile network; reports twice daily; 15 years | Vendor case studies only | Sold via Eurofast (Netherlands-based, with a UK site); no UK installers or projects found |
| Sika Roof Monitoring System (Switzerland) | Water on the vapour barrier, temperature, humidity, plus weather data | On the vapour control layer; 200–300 m² per sensor | Zone only | Vendor says yes | LTE-M / NB-IoT; 10 years of data included | None found | Sika Ireland; understood to be offered in the UK |
| SikaRoof Control (Switzerland) | Conductive layer for electronic leak location | Graphite fleece or stainless mesh under the membrane | The breach, when tested | No | Tested on demand | None found | As above |
| Detec PermaScan (USA/Canada) | Moisture bridging copper tape (C); leakage current through breach (I) | Tape grid on the vapour barrier (C) or on the membrane under overburden (I) | Wet cell ~1.5 m² (C); breach (I) | No | Wired control panel | FM 7745 approval (2022) for PermaScan-C — the only independent performance approval found | None found |
| ProGeo smartex® (Germany) | Electro-impedance across the waterproofing | Contact layer under the membrane | ~1 m² | Point-sensor variant only | Logger on LAN, WLAN or LTE | None found; long reference list (Elbphilharmonie, BMW Welt) | No UK projects listed |
| SMT FutureCast (Canada) | Resistance across moisture tape | Grid on the lowest vapour barrier, or on the membrane (inverted) | Wet cell (conventional); breach (inverted) | No | Gateway to cloud | None found | None found |
| SIHGA Monitorix ROOF IOT (Austria) | Temperature, humidity, water level | Probe in a roof inspection port | The probe | Yes, tool-free | NB-IoT, daily; AA batteries ~3 years | None found | None found |
| FLOscout (Switzerland) | Liquid water at probe base, temperature, humidity | Foam probe pushed into the insulation | The probe | Yes | LoRaWAN, Sigfox, NB-IoT or GSM; 4–10 years | Listed in the Swiss CRB specification library (a specification database, not a performance test) | None found |
Also in the field: Roof Protector (credited to SLD Europe AG in 2018 and to RPM Gebäudemonitoring GmbH, Austria, in 2021–22; current status unconfirmed), roofSec (Austria), Senzomatic (Czech, timber), OmniSense (USA, timber), and drain and gutter monitors such as RoofCare, Blücher and Tremco SmartDrain, which watch the outlets rather than the build-up.
A note on timber decks. NHBC Standards 2026 Chapter 7.1 does not accept CLT panels for flat roof construction on warranted new homes, and BS 6229:2025 says CLT decks should be designed out where possible. That affects every timber-moisture system — Tector, Sensor Innovation, Senzomatic, OmniSense — equally in new-build housing; it does not apply to non-NHBC buildings such as offices, schools and university buildings.
What each does well, and where it falls short
Sensor Innovation into®
Strengths. One platform covers the roof insulation, the drains and the timber. It has the best-documented field data of any point-sensor system: SINTEF monitored 46 and 52 sensors on two Norwegian compact roofs for 16 and 28 months, and sensor positions led to three membrane leaks being found (Asphaug, Geving and Moschetti, Journal of Building Physics 48(3), 2024). Limitations. The warning method in that study was developed with the vendor, the leaks were found by people reading the data rather than by alarms, alarms were off for the first seven to eight months, summer humidity peaks had to be suppressed by rule, and humidity sensors saturated when wet. Point detection only. The 30-year battery claim is untested.
Tector
Strengths. Simple installation by a puck in a cored hole; strong in CLT and mass timber, with UK references. Unusually candid help pages on what drains batteries. Limitations. The sensor sits under the insulation, so water has to pass the membrane, cross the insulation and reach a sensor 20–25 m² apart before anything is reported. The battery claim depends on an active subscription; licence terms renew automatically.
VILPE Sense
Strengths. Dense 4–5 m sensor spacing, an owner-owned-data policy with an API, and the only system that pairs detection with active drying (a sensor-driven roof fan). Limitations. Alerts arrive in days, not hours. Humidity readings can’t always separate a leak from condensation. The bundled SIM is specified for the EU and EEA, and we found no UK installers or projects.
Sika
Strengths. The only membrane manufacturer with a real product, and the only membrane manufacturer’s offer that pairs a warning layer (sensors) with a locating layer (conductive fleece or mesh for electronic leak detection). Limitations. One sensor per 200–300 m² is zone-level warning only. Ten years of bundled data is shorter than most roof guarantees. Worth asking at tender: when the manufacturer’s platform flags a leak, how is the question of cause handled, given the same company stands behind the membrane guarantee?
Detec PermaScan
Strengths. FM 7745 approval for PermaScan-C — the strongest independent credential in the market. Works from construction through service life. Limitations. Construction stage only. Cell-sized location still needs a follow-up survey. North American supply chain; no UK presence found.
ProGeo smartex®
Strengths. The most mature European breach-location grid (since 1992), claiming about 1 m² location, with a long list of landmark projects. Limitations. Must be designed in. All performance claims are the vendor’s. No UK projects listed.
How to choose: seven questions
- Will the waterproofing be covered? If so, the roof benefits from permanent leak detection whatever else is true. Then: breach or wet zone? On a covered roof you open up where it is wet whichever system reported it, so the useful question is how small the reported zone will be, not whether the system points at the hole.
- What is the build-up? EPDM, foil-faced layers, isolated decks and timber decks change what electrical methods can do.
- How big is a zone, and what would it cost to open one up? BS 8102:2022 accepts permanent leak detection on buried decks only where leak locations can be identified and repairability achieved. A zone of several hundred square metres under landscaping is unlikely to achieve that.
- Had the waterproofing passed independent quality assurance before the system was commissioned? Inspection and integrity testing by someone who is not the installer, before cover and again after temporary protection is removed (BS 6229:2025 clause 6.2; NHBC 7.1.11.2). A system commissioned over a leaking membrane records the leak as normal.
- What independent evidence is there? Ask for it. At present the honest answer for almost every system is vendor case studies.
- What happens after year ten? Battery life, mobile network life, subscription terms, data export, and what happens if the supplier goes.
- Who owns the alarm? A named person, written into the O&M manual.
What the UK standards say
- BS 6229:2025 (current since 31 December 2025): a code of practice that recommends integrity testing, not monitoring. Clause 6.1 lists electronic leak integrity testing among the compliance tests to agree; clause 6.2 asks for testing before temporary protection and again after removal.
- NHBC Standards 2026, Chapter 7.1: integrity testing by a suitably qualified surveyor on roofs over 50 m² or hard to access, with a report to NHBC; green roofs tested before covering. No mention of permanent monitoring.
- CIRIA C817 (2024): on warm blue roofs, “a permanent leak detection system … should be considered.” The clearest UK endorsement.
- BS 8102:2022 §6.5 Note 2: accepts warm roof build-ups and loose-laid waterproofing on buried decks where permanent leak detection is employed such that leak locations can be identified and repairability is achieved.
- PCA Buried Roofs guidance (2024): permanent leak detection may be considered where waterproofing is not fully bonded, “but it gives limited identification.”
None of these requires permanent monitoring. All of them ask for integrity testing of the waterproofing, which we treat as part of independent quality assurance: the tester answers to the client, not the installer, and the detection system is commissioned only once the QA has passed.
Related reading
- Start here if you are deciding whether this roof needs leak detection: Should you specify leak detection in this roof?
- Start here if the roof feels high-risk: High-risk flat roof? How to reduce the risk
- Smart roof monitoring: what it is, how it works, and what it cannot do
- How many leak sensors does a roof need?
- How to specify a smart, integrated leak detection system
- Permanent monitoring vs electronic integrity testing
- Electronic leak detection — the complete UK guide
- Leak detection for data centre roofs
Vector reviews roof leak detection proposals as part of a waterproofing design review. Send us the roof section and the system you are being offered, and we will tell you what it will and won’t detect on that build-up. Talk to Vector.