For his Energy Harvesting for Wearable Devices research, mechanical engineering assistant professor Shad Roundy received a three-year $475,806 grant from Analog Devices.

As wearable devices proliferate, their reliance on conventional batteries poses several problems. First, their limited lifetime is an inconvenience to users, and in more critical cases, such as wearable health monitors, may limit compliance. Second, disposal of batteries causes an environmental concern. And, third, device manufacturers prefer a completely sealed device, which is not practical with a battery that must be regularly recharged or replaced. The goal of this project is to develop a wrist worn energy harvester that is capable of 100 uW power generation during normal walking with a total device volume of 1-2 cm3.

Previous research indicates that the upper bound for power generation for wrist wearable systems is approximately 150 uW/cm3. (Of course, this number is highly dependent on an individual’s walking pattern, but his number indicates a practical average upper bound value.) Currently demonstrated wrist worn energy harvesters, both commercial products and research prototypes, generate approximately 10 uW/cm3 or less. All devices, to our knowledge, rely on linear motion coupled to a mechanical dynamic system intended to amplify inertial forces through resonance, or on non-resonant rotational motion. The linear devices have the advantage of amplifying inertial forces, but suffer from displacement limitations and can only be excited by motion along one axis. Rotational devices, such as self-powered watches, can be excited by linear or rotation motion about any axis. However, they do not benefit from inertial force amplification.   By combining the benefits of rotational motion with resonant dynamics, we believe that significantly higher power generation is possible and that the order of magnitude gap between what is possible and what has been demonstrated with existing devices can be closed.

We will address the following major challenges. First, we will evaluate, via simulation, several dynamic system concepts, making use of resonant dynamics and rotational inertia motion. We will use a generic energy transduction damper in these simulations to determine the maximum amount of power that can be generated, and the level of electromechanical coupling necessary. Second, we will evaluate piezoelectric, small scale electromagnetic, and electrostatic transduction methods in order to determine which technology will be able to deliver the needed electromechanical coupling within the form factor required. Third, we will investigate active control over the level of electromechanical coupling allowing for real-time optimization of generated power. The first two activities will result in prototypes having the goal of 20 uW/cm3 power generation while walking during the first year. Along with incremental improvements in years 2 and 3, the third activity will enable further efficiency allowing us to meet the overall goal of 100uW in a volume of 1 to 2 cm3.

Both the use of resonant dynamics in conjunction with rotational systems and active control over the level of electromechanically induced damping are novel aspects of this work which will contribute to the overall state of knowledge on harvesting energy from human body motion.

Learn more about Dr. Roundy on the Integrated Self-Powered Sensing Lab site.