With more than eleven billion Internet of Things-connected devices in the world, and an increasing number of large-scale, distributed networks of sensors, the problem of powering these systems is becoming a pressing issue. When these devices cannot be powered by the electrical grid, the present options that remain leave much to be desired. Rechargeable batteries require a tremendous amount of maintenance to keep them charged, and they must also be replaced periodically. Moreover, these batteries are loaded with toxic chemicals that could cause real environmental harm if we rely on them for the operation of billions of devices.
Energy harvesting is a much more promising option from the standpoint of maintenance and environmental protection, but this option also has some drawbacks. First and foremost, these systems are challenging to design. Whether solar, wind, or other energy sources are utilized, they are often inconsistently available. As such, complex systems are needed to store harvested energy, power up the device only when sufficient amounts of energy have been stored, and make efficient use of the available resources.
While the technology does presently exist to build efficient energy harvesting systems to power sensor networks, selecting the proper components and integrating them into the devices can be extremely challenging. Each device, and method of energy harvesting, comes with its own set of considerations that the power management systems must be finely-tuned to accommodate. This barrier to entry is preventing environmentally-friendly energy solutions from being deployed into the wild, so a team of researchers at MIT developed a design guide for energy management in self-powered sensors to ease the process.
The guide focuses on three topics that are of particular importance to self-powered sensors — enabling a cold start, efficiently converting and storing harvested energy, and control algorithms to make good use of the available energy. To start up cold, the guide details how harvested energy can be stored until it reaches a critical threshold, at which point the device can power up. To keep the system as maintenance-free as possible, batteries are avoided. Instead, the guidelines call for capacitors to store harvested energy until it is needed for operation. Helpful tips are provided to make sure the capacitors can store sufficient energy for system operation, but also to avoid selecting too large of a capacitor which would result in an excessively long charging period. The control algorithms determine when to turn the sensors on and off, and when to take a break to harvest more energy.
As a demonstration of the design principles, the team built a simple device that harvests energy from the magnetic field generated around a wire. This energy was utilized to power a temperature sensor, which could be leveraged to, for example, monitor the temperature of a motor. The collected measurements were transmitted wirelessly to a smartphone via Bluetooth. The design principles are flexible and not constrained to just harvesting magnetic field energy — other power sources like vibrations or sunlight can also be utilized.
The team’s system accounts for many known issues that can be a thorn in the side of developers. Harvesting too much energy, for example, can be a problem for the low-power circuits employed in these types of devices. To avoid minor problems — like exploding devices — the energy management system will automatically adjust the amount of energy being harvested. The guide also takes into account the large amounts of energy that are required for energy-intensive operations like communication.
As a next step, the team plans to work toward more accurate modeling of how much energy comes into the system, and how much it uses during operation. More accurate models could extend the operation of the device, which would allow it to collect more useful data. They also intend to explore less energy-intensive means of communication, like acoustics and optics.