Occupancy and Energy Consumption

 

Buildings are built for people.

It might seem obvious, but the flow of people through a facility is one of the most important and commonly ignored elements of energy culture. Measuring energy consumption in dollars and cents is convenient, which is why foot traffic is often overlooked. But whenever an element is ignored, administrators and managers are paying for it in higher operating expenses, and nowhere is this more apparent than in the case of understanding foot traffic.

At Sapient, our product suite specifically addresses this inconsistency. If our smart outlets collect device-level data, and if the Sapient ML core unites, integrates, and automates the most optimal energy practices, then our occupancy sensors provide a crucial context to the whole solution.

But why spend the time and resources to engineer a state-of-the-art occupancy sensor at all, much less one that uses Bluetooth, infrared, and ultrasound sensors to build out a behavioral model? What context can occupancy provide in energy management?

The answer lies in the behavioral component of energy consumption. For a facility manager monitoring a floor, energy consumption clearly fluctuates throughout the day. What that data cannot tell you is how the energy is being used, why it fluctuates, and whether it can be correlated with a human proxy. If a correlation can be made between power draw and that human proxy, power delivery becomes a candidate for automation.

Occupancy adds a crucial layer of context to the data for managing energy use.

Imagine two rooms full of machinery, one monitored by a Sapient system, and the other by your building’s current BMS console. The correlation of occupancy and energy usage provided by Sapient implies a behavioral dependence to the dollars spent in powering that room. Your BMS, submetering the other room with no occupancy insight, reveals very little information about how that room is used.

Occupancy and power usage are complementary data points. Together they describe why a device is functioning. When coupled with hundreds of people, rooms, and devices, machine learning can use these insights to build a predictive model of energy use. The goal is to isolate negligent or wasteful behaviors and automate the facility’s power delivery to the point of completely eliminating the wasteful consequences of those behaviors. Without occupancy data, there is no way to establish these powerful correlations.

We have not invented occupancy sensing. If you have ever experienced the mild startle of a dark room lighting up the moment you walked inside, you have seen occupancy sensing at work. But you can do much better than settling for the low-hanging fruit that is occupancy at its current complexity.

The implications of knowing how people move and use energy throughout a facility are profound. Think back to those two rooms of machinery, one of which couples occupancy and energy usage data. Both rooms are fluctuating in their energy use, but for some reason, the second does not correlate with typical workday hours. It might simply be that the machinery is not used, and that should be addressed. After all, over $19 billion dollars of electricity is wasted on devices sitting in standby mode every year. Maybe the room is remote enough to discourage its use, or a consistent flow of employees to the room better suit it as a coworking space or conference area than a traditional workplace. With integrated occupancy data, a manager can make judgements, organizational or structural, that take into account a newfound awareness of how people move throughout the building.      

By focusing solely on aggregate power consumption, conventional BMS’s ignore the one element that makes a living, breathing facility live and breath: human behavior. By viewing energy consumption as a function of human behavior, Sapient is bringing energy management into the 21st century.