Not only can wearable devices use vibration energy to charge their own batteries, but they could one day have a big impact on our communities.

That’s the conclusion reached by a group of international researchers experimenting with kinetic energy harvesting. The group – comprised of researchers from the University of New South Wales and the University of Queensland – sees potential to use the method as a valuable tool for things like urban planning and development by detecting people’s modes of transportation.

“Transportation mode detection is important to our communities,” says Sara Khalifa, researcher from Data61|CSIRO, Australia. “It allows researchers to consistently and reliably collect information on individuals’ traveling behavior to inform urban design, real-time journey planning, human activity monitoring, CO2 emissions, targeted advertising and more.”

And that’s a great deal more than just reducing battery drain, as this system can benefit our communities by monitoring for potential hazards such as the spread of disease, providing emergency responders with information for the fastest route to a serious accident and much more.

Kinetic energy harvesting is the process of converting environmental kinetic or vibration energy – wind, waves, vehicle movement, machinery vibration, human motion, etc. – into electrical energy that can be used to power small and low-energy electronics.

The researchers proposed using kinetic energy harvesting’s output voltage as the source for detecting transportation mode. This concept is based on the idea that the vibration energy experienced by a passenger are different depending on the transportation used (walking, running, bus, driving, etc.).

When piezoelectric material – the mechanism this study used to convert kinetic to electric power – is subject to stress from environmental vibrations, it generates an AC voltage proportional to the applied stress. This means different vibration patterns generate different AC voltage patterns, which can yield clues as to transportation mode.


The voltage generated by the wearable records distinctive differences between the user’s transportation mode (see image above). With a data logger, the image below shows a step-by-step diagram of the path to classifying the mode of transport (walking, running, car, train, bus). The researchers were able to achieve 98.84 percent accuracy in determining whether the user was walking/running or in a motorized vehicle. Overall accuracy of classification between motorized vehicles (bus, car, train) was over 85 percent. At the same time, kinetic energy harvesting allowed these wearable detecting devices to stay fully powered.


While this research focuses on detecting the mode of transportation that users are in, the research team continues to explore various other applications such as human activity recognition, step counting, calorie expenditure estimation, user authentication, train route identification and detecting “hot words,” such as “OK Google.”

With this wearable in motion, communities and cities will be one step closer to fully understanding the movement within their walls.

Learn more about kinetic energy harvesting in IEEE Xplore.