The Maintenance Corner
Social Housing ManagementBuilding Pathology

DIY Smart Damp & Mould Monitoring for Housing Providers

7 min read

Have you ever wondered whether you could self-host your own remote data loggers, rather than paying big fees to the suppliers currently on the market? Well, guess what: on paper, it is possible, and it looks a lot easier than you might think.

I want to be upfront. This is not something I have built yet. I am looking into it at the moment and working out what a first trial would look like. So treat this as my planning notes rather than a finished build write-up. If I go ahead, I will report back on what worked and what did not.

The problem

Temperature and humidity monitoring is becoming one of the more useful tools we have for damp and mould. A sensor on the wall gives you weeks of readings instead of a single snapshot on the day of the visit. It helps you tell the difference between a household drying washing on the radiators with the heating off and a property that is genuinely cold and struggling. With Awaab’s Law now live in Scotland, having evidence that a home stays free of damp after the works are done is not a bad thing to have either.

The trouble is how it is sold. The packages I have seen tend to come with a minimum order, often tied to a set number of sensors and a fixed contract term. That is fine if you are ready to roll out across a whole estate. It is far less helpful if you only want a handful of sensors for your problem properties, or to try one in a void before you commit to anything. You can also be left unclear on what happens to the cost once the initial term ends.

What I really want is to buy sensors as I need them, own the data, and not be locked in.

How it would work

Here is the set-up I am looking at. There are four parts to it. I apologise in advance if some of the terms are a bit ‘techy’ but if you don’t understand any terms it’s fine, sometimes you don’t need to know how something works to know that it does work.

1. The sensor. The one I am considering is a small battery-powered wall unit, the ELSYS ERS Lite in its NB-IoT/LTE-M version. NB-IoT and LTE-M are low-power mobile networks built for this kind of device, so there is no Wi-Fi and no hub, and nothing to plug into the tenant’s broadband. On the datasheet it is accurate to ±0.2°C for temperature and ±2% for relative humidity which is more accurate than allot on the market. The manufacturer quotes battery life of up to 10 years, but that depends on the settings and signal, which I come back to below. You set it up by holding your phone against it and using an app over NFC. I am not endorsing it. It is just the option that ticks the most boxes for me so far.

2. The SIM. Each sensor needs a SIM. There are prepaid IoT SIMs where you pay once, roughly £12 for 10 years and 500 MB of data, and you can buy a single SIM rather than a box of 50. A temperature and humidity reading is tiny, so 500 MB should go a long way. The one I am looking at is 1NCE, with Soracom as a backup option.

3. A relay. This sensor does not send data in the format most web tools expect. It sends small encrypted packets over UDP, using either the manufacturer’s own protocol (secured with an AES-128 key that you set yourself) or a device standard called LwM2M. It does not speak MQTT or HTTP. The good news is that the SIM provider offers a free relay that receives those UDP packets and passes them on as a normal secure web request. That is the bit that makes it possible without running your own server.

4. Somewhere to store and view it. The plan would be to point the relay at a small script on Cloudflare (a Worker). It checks and decrypts each reading and saves it to a simple database (Cloudflare D1). Then a private dashboard sits behind a login (Cloudflare Access), so only named staff can see it. At the scale of a small trial, all of this should fit within the free tiers. A humidity threshold email alert could be added on top.

On reporting, the sensor sends a reading every 30 minutes by default, and you can change that over NFC. It can hold up to 50 readings before sending. Sending is what drains the battery most, so one idea I am toying with is taking a reading every 10 minutes and sending them in a batch every 6 hours, which is 36 readings per send. That is not settled. I need to check what it does to battery life before I decide.

On cost, a first trial of one sensor and one SIM looks to be in the region of £80 to £110 ex VAT, plus batteries and shipping. That is an estimate, as UK resellers seem to range from about £66 to £100 for the sensor. Hosting at that scale would be £0.

Where it can trip you up

It looks simple on paper, but there are a few things I need to nail down before I spend anything.

The decoder. As far as I can tell, there is no ready-made decoder for this sensor’s protocol, so a small piece of code would need to be written to turn the packets into readings. I want to confirm the details with the manufacturer first rather than guess.

Signal and battery saving. The 10-year battery figure assumes good signal and a network that supports power-saving mode (PSM). If the network does not support it, or the sensor sits in a poor signal spot, the battery will drain much faster. I need to check which UK network the SIM uses, whether PSM works on it, and what the signal is like at the trial address. A stone-built ground floor flat is not the same as a top floor office.

Buying the SIM. These SIMs are sold to business customers only, so the order needs to go through the organisation rather than a personal account.

Data protection. This is the big one for housing. Readings from a tenant’s home can tell you when someone is in, when the heating is on and how the household lives. That may well be personal data. For a first trial I would keep it to a void property or the office. Before using it in anyone’s home, you would want a disclaimer done and sign-off on where the data is hosted, ideally on the organisation’s own account rather than someone’s personal one. Something I have been toying with is rather than listing property addresses using something that doesn’t identify the property but a different unique identifier which could add that extra layer of security.

Support. If you build it, you support it. There is no helpdesk to ring when a sensor goes quiet. For a handful of sensors that is manageable, for small to medium sized housing associations this could be the trick. For local authorities with hundreds or thousands of loggers, you would want to think hard about who owns it.

Conclusion

I do not think you have to sign up for a big contract just to find out whether monitoring helps. On paper, a sensor, a prepaid SIM and a free-tier dashboard gets you most of the way, at a fraction of the outlay, and you buy as you go.

The plan for a first trial would be:

  1. Confirm the decoder and encryption details with the manufacturer, and the network and battery-saving support with the SIM provider. That costs nothing.
  2. Buy one sensor and one SIM.
  3. Set up the dashboard on the free tiers.
  4. Run it for 4 to 8 weeks in a void or the office.
  5. Report back on the results and the real running cost.

If it works, great. If it does not, I have spent around £100 finding out, rather than committing to a full contract. Either way, I will share what I learn here.

Thanks for reading. If you found this useful, feel free to subscribe via email for new Maintenance Corner posts.