Schneider Electric sought to create a building automation system that would be affordable for small to mid-sized existing properties such as schools and medical buildings, and that could be installed without causing building downtime or business disruption. The resulting product is the SmartStruxure Lite building automation solution, an efficient way to future-fit small- to medium-sized buildings to control HVAC, lighting, and metering, and save energy and time, and improve comfort with minimal impact on operations.

To meet the objectives, Schneider Electric’s design team recognized that the system would have to utilize wireless and web technologies to minimize up front hardware and labor costs, such as installation of cabling. To control the myriad devices connected through the wireless mesh, Schneider Electric created the Multi-Purpose Manager (MPM), a programmable controller, gateway, Web server, and protocol translator. In effect, it’s a classic case of an IoT Gateway providing remote control over various end points in the building.

The MPM, within its SmartStruxure Lite building automation system, requires management of considerable variety and amount of data. For example, communication and routing data, control variables and history, system status, state data for networked equipment, raw sensor data, and aggregated/processed sensor data, are all collected. Therefore, the design team recognized early on that a database management system would be required on the gateway.

To meet the objectives of an affordable building automation system, the MPM would need to minimize hardware and maintenance costs, so components with a relatively short lifespan, such as hard-disk or solid-state drives, would not be included. Consequently, an in-memory database was the only option.

But not all in-memory databases are created equal. This database would need to run entirely from memory, i.e. not bootstrapped from an on-disk image, no transaction logs, etc. To further minimize component costs, the in-memory database would have to use available memory efficiently, and have a small code size.

After evaluating several commercial and open source database systems, eXtremeDB was selected because it met these requirements at an attractive price point. eXtremeDB is similar to NoSQL databases in that it offers a direct, native programming interface. But eXtremeDB lets the database designer define the database schema.

This fit well with SmartStruxure Lite because the data requirements are well-understood and a database system that understands the structure of the data under management can manage that data more intelligently than opaque data, as is the case for “document” or “key-value pair” database systems.

Steven Graves co-founded McObject in 2001. As the company’s president and CEO, he has both spearheaded McObject’s growth and helped the company attain its goal of providing embedded database technology that makes embedded systems smarter, more reliable, and more cost-effective to develop and maintain. Graves is an advisory board member for the University of Washington’s certificate program in Embedded and Real Time Systems Programming.