Vibration sensors used to collect structural information

Hae Young Noh, a Carnegie Mellon civil and environmental engineering assistant professor, recently won the NSF CAREER Grant for her research on “Structures as Sensors.” Her research uses small cylindrical sensors located at strategic positions to measure vibrations generated by human activity, then extrapolates information about their quality of life.

She states that each person’s gait is distinct enough to be used as identifiers. “Each step,” she explains, “is like a mini earthquake with its own epicenter.” The gait of a person can be characterized by various parameters such as balance, stability, strength, and speed. The vibrations created by steps hold more information such as location, identity, level of activity, health status, and even mood. Although she is still fine-tuning the data related to mood, Noh argues that mood is very certainly reflected in one’s gait. “A depressed person walks very differently from one that isn’t,” she remarked.
Vibrations are also generated by machines and other household objects. For example, when someone turns on a tap, the water running through the pipes generates a characteristic vibration. Thus, these sensors receive a lot of information about many different actions occurring within a building. Processing this data is, in Noh’s opinion, one of the most challenging aspects of the project. She said, “The challenge lies in the data analysis and the signal processing. It is a very convoluted signal, giving us information about people, about structures and about the environment. Now that we have all this data, how do we take what we want? That is something I’m working on.”

Noh explained that, at a very basic level, the data works like an identification match. One can look for a “gait match” in a database, and if the match doesn’t exist then the person is unauthorized in the building. This has several different targeted applications such as security. Another application would be in the retail industry where it could be used for customer behavior analysis. Furthermore, it could give information about road traffic and the climate. Another application for this technology would be to get data about various structures. For instance, when installed in railway structures it could be used to get data about the faults in the tracks as the vibrations change.

This technology has been developed keeping the specific needs of nursing homes for the elderly in mind. Noh works with the Vincentian Nursing Home and the Baptist Nursing Homes in Pittsburgh, so she explained, “there is someone whose gait suggests that the quality of their walk is deteriorating and they should probably slow down and take rest. In that case, the system would alert a caretaker who could handle the situation. Thus, the person in question doesn’t have to completely be confined to a wheelchair to prevent falls in case there aren’t enough caretakers to shadow them all the time.”

Noh’s research is particularly advantageous because of its noninvasive nature. Cameras and microphones record very personal and sensitive data by nature, which causes several people to reject them as a method of information gathering. Secondly, wearables such as smartwatches require people to place a device on themselves, which could be seen as an effort on the wearers’ part. Noh reflects that this sensing system requires minimum interaction from the users with the system itself. Regarding this project, Noh predicts that the commercialization of this technology would take five to six years, and the technology for advanced mood-based and medical analysis would take a little longer.

Noh said, “It is such a convenient way of discovering so much. It is seemingly impossible in the beginning, when one sees all the information and the noise stored in one signal. But, then it all comes together and makes so much sense.” While reflecting on this technology in improving quality of life, Noh explained, “this research has so much of a human-aspect to it. It is very rewarding to work with people in such a tangible and direct manner.”