A new structural-vibration sensor system inspired by scorpions’ predatory habits detects building occupants by their walking patterns. Walking vibrations produce a unique signature pinned to individual occupants.
“It helps with building-energy management and to control the HVAC systems, based on occupancy,” says Hae Young Noh, assistant professor of civil and environmental engineering at Carnegie Mellon University. The practical application of Noh and her team’s work goes beyond climate control. Noh’s system also can be used for real-time security as well as jobsite safety and productivity monitoring.
The system collects data by affixing accelerometers and geophones to a structure. These vibration-measuring devices look for patterns.
“Amplitude increases as you walk toward the sensor and decreases as you walk away,” says Noh. Depending on a person’s weight, gait and physical health, the vibrations can identify who is approaching.
“We did an experiment and had 96% accuracy in identifying people,” says Noh, whose background is in structural monitoring. She says the traditional approach to gathering structural data is direct. “If we want the temperature in a room, we put in a thermometer. If we want occupancy, we put in an occupancy sensor,” she says.
But the deployment and upkeep of these many instruments is expensive and time-consuming, she says. So Noh and her colleagues sought an approach that is easier and less costly to maintain.
“We came up with the idea of using indirect sensing,” she says. This means non-dedicated, passive sensors, installed throughout a structure, monitor frame vibrations and determine multiple things at once, such as building temperature and environmental events. It puts the majority of research time and expense into processing the gathered data instead of deployment and upkeep of the devices, says Noh.
The idea was born from a conversation with a colleague about how scorpions hunt, Noh adds, saying, “The scorpions feel the vibrations of the prey from the ground, and they follow those and catch them through the vibrations.”
With this method in mind, Noh found research by a Caltech professor who looked at the pattern of vibrations in a building and found repeated profiles that were related to human activity. “All the motions came up, all the traffic outside, the heating-and-cooling system but also repeated profiles from the people,” says Noh.
She started by detecting the damage in the structure and made the structure itself user-aware and then environmentally-aware. “The sensors enable human and building interaction,” says Noh. For example, if the sensors are applied in a nursing home and someone falls in their room, the vibrations of the fall would alert staff.
Noh’s prototype system validated the concept through experiments. Now, she and her team are working with companies such as Google and Intel, among others, for possible large-scale deployment.