Retrofit programmes funded under SHDF are under pressure to prove that fabric and services upgrades actually deliver warmer, healthier, cheaper-to-run homes. That means a plan for IoT sensors that is credible to funders, fair to residents, and practical for site teams under tight access windows. Choosing the right mix is not about gadget count; it’s about selecting sensors and connectivity that map cleanly to the outcomes you’ve promised and the constraints of occupied homes.
TL;DR
/>
– Start with outcomes: damp/mould risk, comfort, and energy flows, then select the smallest set of sensors that can evidence each claim.
– Design for occupied homes: unobtrusive devices, long battery life, tamper-aware mounting, and a connectivity plan that works in solid-walled blocks.
– Contract for data, not dashboards: insist on open APIs, export rights, and end-of-programme data delivery in a usable format.
– Sequence installation with the works: baseline before measures, protect devices through messy phases, and commission after services are live.
– Measure value with simple, comparable KPIs: normalised energy, hours in comfort band, and mould risk flags tied to remedial actions.
Getting the brief right: specifying sensors to evidence SHDF retrofit outcomes
/> Begin with the outcomes your scheme must demonstrate: reduced excess cold, lowered mould risk, improved comfort, and energy performance aligned with your chosen measures. From there, choose sensors that provide clear, comparable evidence. Internal temperature and relative humidity are the bedrock for comfort and mould risk. Add CO2 for ventilation effectiveness where you’re altering airtightness. For fabric performance claims, you may specify surface temperature probes on known cold-bridge locations or short-term heat flux plates on a sampling basis. For systems, consider electricity sub-metering for heat pumps, flow/return temperature probes, and, if applicable, pulse read heads for legacy meters.
Connectivity is the second pillar of the brief. Wi‑Fi is unreliable in many social homes and burdens residents. LoRaWAN or cellular IoT provides a contractor-controlled network that doesn’t rely on the occupant. In dense blocks or solid masonry, a small number of gateways per stair core may be needed; in scattered street properties, battery cellular units can avoid gateway installs altogether. Spell out minimum battery life under your chosen transmission interval, device ingress protection where moisture is likely, and mounting requirements that minimise damage and discourage removal.
Data rights and interoperability need explicit clauses. Require open, documented APIs and the right to export raw and processed datasets during and after the project. Agree time-stamped data retention and hosting location in line with your organisation’s policies, and set out who owns derived analytics. Make room in the spec for basic security practices: unique device credentials, firmware update approach, and clear responsibility for vulnerability patches. In parallel, agree a proportionate privacy approach that keeps personally identifiable data out of the payloads and treats occupancy signals (like CO2) with care.
Finally, write the brief to be installable. State acceptable dimensions, low-profile fixtures, tamper evidence, clear labelling, and QR-coded IDs tied to a flat-level asset register. If you’re sampling, define the sampling logic up front (by archetype, floor level, exposure) so it survives tendering without dispute.
Interfaces and risk on occupied schemes: installation, data and tenancy management
/> Most SHDF schemes live and die on access and sequencing. Sensors must be in before messier trades, protected while EWI or internal wall insulation goes on, and recommissioned when services are live. That means the IoT supplier has to be treated like any other subcontractor: RAMS, inductions, programme slots, and snagging. The retrofit coordinator or client-side technical lead should link sensor locations to the risk assessment — thermal bridges, past damp reports, and rooms receiving ventilation tweaks. Resident liaison officers should set expectations early: what goes on the wall, how often it talks, how data is handled, and how to report issues.
# A mid-rise SHDF scenario under programme pressure
/> A 1960s mid-rise block in the Midlands is in a Wave-funded programme. The main contractor has EWI due to start on Monday, while the retrofit coordinator needs a two-week baseline from 30 flats across all orientations. The IoT supplier arrives with LoRaWAN devices, only to find the stair cores soak signal and the agreed comms cabinet is locked. The housing officer is juggling access and a handful of refusals, and the M&E subcontractor is swapping in demand-controlled extract fans the same week. A last-minute RFI flags that gateway power requires a spur from a communal board that’s not been scoped. The site manager wants to push ahead, but the coordinator knows a broken baseline will weaken the evidence pack to funders.
# Common mistakes
/>
– Installing after major measures: skipping a pre-works baseline makes it hard to show improvement rather than weather effects.
– Picking Wi‑Fi devices to save capex: reliance on resident routers often leads to holes in the dataset and complaints about bandwidth.
– Over-specifying sensors: too many device types inflate install time and confuse maintenance; stick to what proves outcomes.
– Forgetting the removal plan: end-of-programme retrieval or handover to BAU is often omitted, leaving orphaned kit and data silos.
# Checklist for SHDF-ready sensor procurement
/>
– Define outcomes and map each to the minimum viable sensor set and sampling strategy by archetype.
– Specify connectivity with a site survey plan; include gateway power, mounting, and backhaul options where signal is weak.
– Lock down data rights: API access, export formats, time-stamped data retention, and end-of-contract data delivery.
– Build an install sequence tied to the construction programme, including baseline, protection during works, and recommissioning.
– Agree a privacy and consent approach with resident liaison, including opt-out handling and plain-language notices.
– Set QA processes: device ID register by dwelling, calibration certificates or tolerance statements, and commissioning check records.
Risk sits at the interfaces. The client IT team must sign off data handling before kit lands on site. The principal contractor controls access and safety, so the IoT supplier needs a named interface and timeslots that don’t clash with demolition, drilling, or wet trades. The legal team should underpin consent forms and opt-out routes that don’t jeopardise the sample. And someone has to own replacements if devices are dislodged during works — spell it out in the subcontract so it doesn’t surface as a variation mid-programme.
Proving value: from raw sensor feeds to decisions and programme reporting
/> Value is measured in a handful of understandable KPIs. For comfort, track hours spent within agreed temperature bands by room and archetype, and surface where residents are under-heating after retrofit. For damp and mould, combine temperature and humidity into a simple risk index and log how quickly alerts translate into visits and fixes. For energy performance, compare pre/post electricity or gas consumption for system upgrades, normalised for weather using standard degree-day methods and mindful of occupancy. Keep sampling consistent and be transparent about gaps in the dataset; short outages are normal on live estates.
Turn data into decisions, not just dashboards. Weekly exception reporting to site and housing teams should flag flats at risk, missing devices, and persistent comms dead zones. Tie alerts to an action log so you can show that risk was managed, not just measured. For programme reporting, focus on comparables by archetype and phase rather than big headline claims. Housing teams will value fewer repeat damp visits and quicker identification of ventilation issues; funders will want to see trends that align with your measures.
Commercially, judge offers on total cost of ownership. Cheap devices with proprietary portals can be costly if they lock you into licences beyond the project or make data extraction painful. Budget for gateway power, SIMs or backhaul, and site time for install and retrieval. Ask bidders to price replacements and churn explicitly. And insist on a clean handover pack at the end: asset register, floor plans with device IDs, commissioning logs, API keys, and an agreed final data export that can live in your data lake.
What to watch next: local authorities and housing providers are increasingly aligning sensor choices with damp and mould action plans and planned heat pump rollouts. Expect greater scrutiny of data portability across phases, and more emphasis on small, robust devices that can survive busy retrofit sites.
FAQ
/>
Which sensors are typically used to evidence performance on SHDF retrofits?
Most schemes use internal temperature and relative humidity sensors in key rooms to track comfort and damp risk. Many add CO2 to assess ventilation where airtightness or extraction is changed. For systems, electricity sub-metering for heat pumps and simple flow/return probes are common, with targeted surface temperature or short-term heat flux measurements to evidence fabric improvement.
# How do we handle data privacy and resident consent in occupied homes?
/> Keep payloads lean and avoid collecting personally identifiable information. Provide clear, plain-language notices explaining what is measured, why, and how long it’s kept, along with an easy opt-out route that doesn’t penalise the tenant. Work with resident liaison teams to plan visits and handle concerns, and complete a proportionate privacy assessment that your organisation is comfortable with.
# LoRaWAN or cellular: what works better for blocks and street properties?
/> In mid- and high-rise blocks with thick walls, LoRaWAN with a few well-placed gateways can provide reliable coverage without entering every flat for network changes. For dispersed street properties, battery cellular devices can remove the need for gateways altogether. Site surveys matter: stair cores, lift shafts and plant rooms affect coverage, and you’ll need to plan power and mounting for any gateways you do install.
# Who should install and maintain the sensors during a retrofit programme?
/> Treat the IoT supplier like any other specialist subcontractor, with RAMS, inductions and a named interface to the principal contractor. They can install during planned access windows, often alongside the M&E or ventilation team, with the retrofit coordinator verifying locations. Agree up front who replaces devices if they’re damaged during works, and whether kit is retrieved or left in place for BAU monitoring after completion.
# How do we prove improvement when weather and occupancy change over time?
/> Use a short pre-works baseline and compare like-for-like periods, adjusting energy data with standard weather normalisation methods. Stick to a consistent sample and be explicit about gaps or refusals. Pair sensor data with simple operational evidence — reduced damp interventions, fewer comfort complaints — to build a credible picture rather than relying on a single metric.
