Over the past few years, the potential of IoT devices to help housing providers improve their services has become clearer. From improving a property’s sustainability, reducing repair costs through proactive maintenance and improving safety and security, and as IoT technologies get better and the barriers to entry become lower, it’s now clear that the deployment of IoT devices has many practical and value-adding use-cases.
If we ignore the hype surrounding data, IoT devices, and digital twins, the reality is that it’s difficult to launch initiatives using these technologies. For example, creating a digital twin of your assets backed with real data to aid decision-making through IoT devices in all properties is an admirable goal, but much further away than many realise and an enormous undertaking.
These systems are advanced, expensive, sometimes incompatible with current systems and require a lot of effort to build. But although these are long-term goals, that’s not to say we can’t begin making progress today – but how do we get there?
Small roll-outs of IoT devices targeting specific problems are a good place to start as many people and organisations can be wary about large-scale deployments of new and untested technologies. Choosing one challenge to solve within a small set of homes is a great way to prove the concept of IoT devices and paves the way to make the business case for a larger roll-out, eventually feeding into a digital twin.
A relatively simple example could be rolling out temperature and humidity sensors to a block of flats in order to understand the flats’ propensity for mould and damp; it’s then possible to use that data to make decisions about how best to intervene and help the tenants.
Software and hardware considerations
We’re all now discovering more about IoT devices and the challenges of deploying them. Hardware and software roll-outs of devices across many buildings pose many issues: how can we remotely manage all these IoT devices; how can we parse all this data; and how can we make better decisions from this data?
These technical challenges sit alongside concerns associated with how tenants will react to having devices placed in their homes: who ‘owns’ the information gathered by IoT devices; how is the data secured; what is the data used for; and what happens when the devices break?
Choosing suitable hardware is essential; these are some of the factors that we have had to consider and the lessons we’ve learned:
- While a device such as a Raspberry Pi seems like a great entry point, they exist primarily as development units. Developing a Raspberry Pi or an Arduino into something you can guarantee to be safe, secure and professional is hard. The gap between a production-ready unit and the development kit is considerable so picking pre-built devices from a hardware manufacturer (instead of building in-house) is strongly recommended.
- A useful option for sourcing hardware is partnering with research groups or universities. Similar to housing providers, there is significant interest across academia to see how IoT devices can be safely deployed, and many of these groups are very interested in being involved with external partners for real-world insights.
- Many of the problems that tenants might have when installing IoT devices in their homes, such as connecting to wifi, changing batteries and gathering data, can be solved with technology. We’ve seen this in other sectors (such as manufacturing) where the ideal device is wireless and battery-powered, and therefore causes as little disruption as possible. We’re fond of Monnit devices; these are cheap IoT sensors that communicate data to the cloud over 3G, are all wireless and have very long battery lives.
- Battery life is an extremely important consideration; the overall financial difference between a three- and 12-month battery life is enormous. Often, by the time an engineering team has changed a sensor battery once, all of its financial benefits have been lost.
As wariness around adopting new technologies is perfectly normal and will inevitably cause barriers to progress, it can be mostly mitigated by using sensors with which tenants need have no interaction.
The chosen software stack is also important, with many options available:
- The lightest touch option will always be the proprietary software that might come with the hardware you’ve bought. Many sensor manufacturers also offer a cloud platform for viewing and interacting with your IoT sensors’ data, such as Monnit’s iMonnit platform.
- Microsoft offers the Azure IoT Hub, enabling you to monitor the health of your IoT devices and receive their data, which in turn can then be fed into your other systems. IoT Hub is a powerful, industry-standard way of looking after a fleet of IoT devices.
- Capturing data in IoT Hub lets you use Microsoft’s set of tools to handle streaming data. Whole articles could be written about this pipeline, but Azure Functions and Streaming Analytics can parse and analyse your data in real time, and these insights can then be pushed into any of your existing BI solutions. Passing data ‘outside’ like this enables it to be mixed with other data you have, such as maintenance jobs, scheduling and tenant satisfaction.
By focusing on the benefits for tenants while incentivising usage and minimising the tenants’ responsibilities, you can overcome their natural resistance to new technologies.
Small roll-outs to prove the concept of IoT sensors can be a powerful way to show their value; for example, targeting a single, common issue such as damp and mould is a good way to start. Modern IoT devices provide cheap and reliable ways of gathering data, and mixing these with Microsoft’s technology stack provides a trustworthy base to make data-driven decisions.
The smart home revolution is here, and housing providers have a significant opportunity to embrace this technology in a way that requires little input from tenants, but can ultimately help them as much as meeting housing providers’ own needs. It might be an adventure to accomplish this but it will be worth it in the end.
Alex Bookless is the technical director and Andrew Balance is a data scientist at Waterstons.