The number of start-up companies selling sensors is growing rapidly. Sensor company, Spaceti just won the MIPIM PropTech start-up competition in Cannes and the Scrabble-tile sensors of Disruptive Technologies continue to catch the eye. There is good reason for this popularity. The visible data intelligence we have about how properties operate is poor. Cloud-based data insight on operational performance is a fledgling area of technology. Everyone knows though that the creation of digital assets for real estate is a huge opportunity. This is why IBM, Microsoft and Google are all interested in this opportunity. These tech companies all have powerful data analytics platform as well as Machine Learning and AI capabilities. However, they are only powerful if they have good quality and plentiful data, and that's the problem: the data infrastructure does not exist for real estate in many cases.

Everyone knows though that the creation of digital assets for real estate is a huge opportunity. This is why IBM, Microsoft and Google are all interested in this opportunity.

Data infrastructure

There are already two, well-developed areas of data infrastructure. The first is the Building Management System (BMS), which is an internal network in a building which connects all of the existing sensors and HVAC equipment for automated control. There are other internal building systems, like access control and lift control systems, which usually operate in parallel. The second area of data infrastructure is the Cloud data infrastructure for storing, structuring and processing large quantities of data. These two areas of internal building systems and external Cloud infrastructure are not usually connected, although creating this connection efficiently is what Demand Logic does.

New sensor networks

In many properties, we are now adding a third segment, which are new sensor networks. Often added to a property as a parallel network connected to the Cloud for a specific use case, i.e. occupancy or wellbeing or equipment monitoring, without relating it back to the existing building systems which control the property operations. If there is a disconnection, then the wellbeing sensors could tell you that there is an air quality issue, but not how to change the HVAC operation to improve the internal environmental conditions for people.

In our conversations with property and facilities managers, the Demand Logic team is often asked whether we can add in a range of new IoT sensors. We can, but the reason why new sensors are required is not always clear. I always encourage our customers and partners to think first about what data exists in their properties. On the BMS for example, you will always find temperature sensors and getting the internal temperature of an office under control is an essential first step for wellbeing. If you add in new air quality sensors and your BMS and HVAC is not controlling the temperature well, then there is little chance you will be able to successfully effect an improvement in air quality because the BMS and HVAC also control the air flow.

Demand Logic has recently started to use sensors to supplement the existing BMS data, which include wellbeing sensors; data about the condition and run-time of the HVAC equipment; utility meters; and occassionally lighting data. The way in which we are using new sensors is for specific use cases: more details about wellbeing, particularly CO2 levels; vibration analysis for monitoring the condition of critical plant; water leak detection to reduce insurance risk; water temperature for health & safety risks associated with the breeding of Legionnella bacteria; and occupancy for better space planning and meeting room management. By choosing the right sensor manufacturer we are able to address all of these use cases by installing a single wireless sensor network connected to our Data Acquisition Device (DAD). This means that the security risk is concentrated in a single pathway for the data rather than multiple external connections for each set of sensors for each use case. It also means that the data can be viewed together in one place and, for instance, the occupancy data can be used to optimise the HVAC control via the BMS.

Understanding the data you have

Before we recommend new sensors, we always ask our customers and partners what data they already have. I was recently asked if Demand Logic could monitor the lifts in a property because the FM felt like they were experiencing a lot of faults causing disruption to the employees. Now we could, for instance add wireless vibration sensors on the motors to predict a failure, but we didn't know what the most common failures were that the FM was experiencing. Nor did the FM and they hadn't asked the lift maintenance company. The first step though was to find out what were the most common failture modes. to see whether the addition of new sensors would offer the right cost-benefit. The lift control system also had the intelligence to direct the build users to the next available lift going to their floor. My next question was whether this data has been logged, trended and analysed to see whether any of the four lifts were experiencing more usage and wear & tear than the others. Did the usage correlate at all with the faults? Again, this existing source of data had not been looked at.

I've had similar conversation about occupancy data. Can we use sensors to monitor whether spaces are being used and how many people are in those spaces? We can, but this can be expensive. I've been asked this for properties where these is an access control system, both at the front floor and one each floor. Yet, the data from this access control system has not been analysed to see what the total occupancy numbers are throughout the day. Even for properties which are expected to run 24/7. Now, more granular data on occupancy and usage patterns from new sensors can have more value, but I would still recommend starting with the data you have.

More granular data on occupancy and usage patterns from new sensors can have more value, but I would still recommend starting with the data you have.

Adding new sensors to create more data can be a good idea. It is only a good idea if you can make use of that data. Does the new data create value for the property? Can you use and act on existing data before adding new data sources? If you add new sensors will they just tell you that you have a problem or will they provide actionable intelligence so that you can improve the performance of a property? Mapping the existing sources of data, from building systems, against the customer use cases will help you identify the data gaps for new sensors and also evaluate the cost-benefit analysis of integrating or adding new sources of data to deliver sustained benefits for the property and its occupants.