Power supply is a bugbear of the outdoor IoT, especially remote and distributed sensing or monitoring applications. Mains electricity supply is usually not available, while batteries involve costs and logistical challenges, or constrain utility. In monitoring of gas and water supply grids, the need to extend battery life precludes regular wireless transmission of sensor readings, which in turn impairs their value.
Wireless real time communication with control centers is ideally required every 2-3 minutes for optimum active pressure management in gas networks, for example, and yet if monitoring is supported by batteries the costs and logistics of regular replacement in the field restrict such data collection to once a day in practice. This constraint applies more generally to IoT beacon systems spanning many applications such as proximity management, where the cost of battery replacement has to be balanced against value of frequent data updates.
For this reason, there has been growing interest and investment in energy harvesting, exploiting either power available from the system being monitored itself, or from the ambient environment. The underlying principle is that there is plenty of energy around capable of powering IoT devices in the vicinity, with the challenge being to capture it efficiently.
In some cases, as in the case of gas or water transmission, the underlying source of energy is in the commodity being transmitted, with scope to exploit transient rises and falls in pressure. In other cases there may scope for harvesting some energy source of anthropogenic origin, such as the electromagnetic waves emanating from electricity power lines, or the vibrations of a bridge. In yet other cases energy may have to be harvested from the environment, perhaps exploiting temperature differences or natural seismic vibrations, or indeed light.
Energy harvesting on a macro scale, as we know, has a long history dating back centuries to wind and water mills where natural energy was converted directly into grinding work. Much more recently there has been the huge uptake in generation of sustainable energy from wind and solar power, but in the case of the IoT we are talking about reducing this to a small scale. This requires development of dedicated silicon for processing the energy capture and also materials that optimize that process as well as cost.
Technologies that harvest energy on this scale fall broadly into two camps, piezoelectric generation and electromagnetic induction. There are now working commercial examples of both, each of which can overlap but have different pros and cons.
These are easily confused, both harvesting the energy of vibrations, which being cyclical can generate electricity by the same underlying principle as the electric motor. The piezoelectric effect feeds off tiny vibrations in a material resulting from mechanical stress, which in the case of small IoT devices would be imposed on a membrane or cantilevered beam that is integral to the system. The mechanical energy is converted, or transduced, into electric current in certain materials and then captured via an attached dedicated chip that powers the integrated circuit of the device.
This might seem like the illusory or hypothetical perpetual motion machine that goes on for ever in the absence of an energy source, contravening laws of thermodynamics. But that is not the case, because energy is input during manufacture or configuration in clamping the material. This energy is released over the lifetime of the device as vibrations feed it back continuously. This would not last for ever because the process would eventually yield to material fatigue, but with suitable choice this would not happen for years, decades or even centuries, far longer than any battery.
A point to emphasize is that, compared to wind and solar energy conversion, which can generate huge amounts of power in watts, piezoelectric materials usually generate much less energy measured in milliwatts or even microwatts. Their advantage lies in the persistence of the ambient vibrations they exploit.
Electromagnetic harvesting also derives power from vibrations in materials that may result from being trapped, or else as springs that are constantly compressing and expanding after initial launch. In this case, however, electromagnetic induction converts mechanical energy into electricity, the same principle used in transformers and induction hobs for cooking.
The principle is that electric current is induced either when an electric conductor is forced to move through an electric field, or when a magnetic field surrounding the conductor changes continuously. In the IoT case, the latter usually applies as he system comprises tiny magnets attached to a source of mechanical movement, along with coils in which the electric current is then induced. Again, a dedicated circuit would then pick up and feed this current to the IC of the IoT device. At this scale, the electromagnetic harvesting is an example of MEMS (micro electro-mechanical system), which became possible with modified semiconductor device fabrication technologies.
Belgium-based e-peas is one firm that makes energy harvesting semiconductor products that can operate with both the piezo and electromagnetic methods. It collaborated with chip and module maker Sequans to develop a cellular-based LTE-M and NB-IoT connectivity solution shown at the recent virtual Consumer Electronics Show (CES) 2021.
Another player is the Solar Impulse Foundation (SIF), based in Lausanne, Switzerland, and best-known publicly for completing the first solar powered flight arounds the globe in July 2016. It has developed coin-sized microturbines for energy harvesting in gas and water distribution pipelines, exploiting those fluctuations to generate electricity off the grid. The name simply refers to a small turbine similar in concept to those that derive power from water or gas consumption on a macro scale. SIF has developed two devices, one outputting 30W and the other 150W, considerably more than the piezo systems.
Indeed, one limitation of piezo generators is their limited power output which confines them to very low energy devices. Nonetheless, there are plenty of these and there is growing interest and research in nanoscale piezo electric generation for applications in healthcare, with potential for powering biomechanical implants.
Energy harvesting at small scale is then is a fertile field of investment and research, being driven by the IoT across diverse sectors and scales.