I didn’t start this week’s newsletter planning to write two stories on smart building systems, but here we are. A few weeks ago, a startup called Mapped launched and announced the $3 million in funding it had raised back in February of 2020. Mapped’s pitch was that it would provide an API for building data, no matter what proprietary systems the building used.
Smarter buildings are the future, and having a digital version of an office or shop offers numerous benefits, among them saving energy, tracking airflow for improved ventilation, and better public safety. In a smart building, during an emergency, it could be possible to track employees in real-time through their badges, letting first responders quickly see where they need to focus their rescue efforts.
But if you want to digitize a building, you have to map it first. And to map a building, even if you plan on using an open source data schema such as Brick (which I covered in the main story), you still have to somehow find all of your fire alarms, HVAC chillers, thermostats, lights, etc., and label them. Yes, the facilities team will have some idea of what’s in a building and even where it is in the building, but buildings change as people bring in new systems and devices.
Mapped helps make the process of figuring out what’s in your building and the relationship those items have with each other easier. It does this in two ways.
First, the company makes a physical device that plugs into a variety of systems and starts ingesting the traffic data from them. It sends that data to the cloud and analyzes it to understand what types of devices are in the building and what those devices talk to. Mapped also incorporated cloud-based data, such as information coming from guest access systems or 3-D renderings.
Second, it tries to build models of individual devices, including what they are designed to do, by using natural language processing to “read” their manuals and figure out how they handle data. So far it has models associated with 30,000 devices that it has built using this method.
After ingesting the data, Mapped’s device analyzes it to provide a map of what devices are on the network and what they talk to, and then makes that accessible to developers through an API. So if the data is BACnet, Modbus, IFC, or any other format, Mapped pulls it in and spits it out in a unified format that developers can work with. Mapped also lets developers set limits on what systems can talk to other systems.
A company might build an app using the Mapped data that ties the guest management system to the Wi-Fi network so guests get a Wi-Fi login when they enter the building, but it might limit their access to everything else. Or it might limit their credentials to certain floors of the building, so they can’t even access them through the elevators.
Building data is big business for the tech world. Microsoft launched a partnership with Willow and RealEstateCore last year to create digital twins of offices, and Google is trying to model its own buildings and launch its Digital Building Products service. Siemens signed a partnership last year with Salesforce linking employees with HVAC and lighting controls for a touch-free office.
Later on, the plan is to introduce applications that can take advantage of this data (currently a customer has to write their own). Those applications can tie devices on one system to devices on another system in ways that aren’t really an option today. For example, the building system might be able to learn when a meeting is scheduled and how many attendees are expected from a Microsoft 365 entry, then trigger the HVAC in that room to lower the temperature before the temperature gets too warm and triggers the thermostat.
That’s a really simple example, but as companies start to look at their buildings holistically, they may also see patterns tied to their employees or their operations that can lead to new insights, allowing them to save money, energy — even boost productivity. But before we can do anything like that, building managers have to map out what they have.