Who’s Got the Power: Energy Harvesting Ecosystem Seeing More Design Wins

Increasing concern for IoT sustainability and, more practically, for greater sensor device autonomy has helped drive extensive R&D, but few concrete applications, over the past decade in the field of energy harvesting (gathering minute quantities of energy from a device’s surroundings and using this to charge a battery or supercapacitor). Technologies go from the now-mundane, such as photovoltaic harvesting, which is already widely deployed in outdoors IoT installations, to the quirky, such as harvesting energy from sweat for applications in wearables, with these latter projects mostly in early research stages. When referring to energy harvesting, this ABI Insight refers specifically to non-photovoltaic methods of harvesting energy—and primarily vibration and RF harvesting. When the futuristic is peeled away from the feasible, a number of recent market developments have demonstrated industry interest, pointing to promising new applications that could transform how the IoT is powered. These include:

• In August 2020, leading industrial equipment manufacturer Hitachi announced that it was acquiring vibration energy harvesting company Perpetuum to leverage the latter’s strengths in train maintenance using self-powered sensors.
• In September 2020, electronics manufacturer SMK Electronics (manufacturing IoT communication modules) partnered with Atmosic, a company pioneering RF energy harvesting. Atmosic was named in 2019 by ABI Research as a Hot Tech Innovator in wireless connectivity.
• Strong funding rounds were reported by several other energy harvesting startups. Another ABI Research Hot Tech Innovator, Wiliot, in 1Q 2020 received US$20 million and industry backing (Maersk, PepsiCo, Verizon, NTT Docomo). e-peas added to its 2018 partnership victory with Fujitsu to raise EUR€8 million in its latest 2020 funding round. Everactive raised US$35 million in its latest funding round, and has previously received industry interest from sensor and equipment manufacturers, such as ABB.

There is no lack of energy for discussing power consumption in the IoT space. One of the principal stumbling blocks facing those wanting to deploy an IoT solution—and consequently those manufacturing battery-powered IoT devices—is the power consumption of a sensor node or device. A short battery life means changing or re-charging batteries more frequently, resulting in high recurring costs in material and labor, as well as significant logistical barriers, especially at high device deployment volumes. A larger battery means bigger form factor and higher Bill of Materials (BOM). As a result, focus has gone into reducing power consumption for devices; device management, connectivity specifications (such as BLE or LPWA networks), and MCU power savings have been at the forefront of the battle. Energy harvesting is increasingly coming into its element with these developments, which enables a suitably low current draw for each transmission to allow a device to be powered by the minute quantities of energy generated, measured in milliwatts. While photovoltaic harvesting has seen adoption, applications are very specific and come with drawbacks to universal adoption—namely fragility, the need to be exposed directly to sunlight, high cost, and impact on device size. Further technologies such as vibration or RF harvesting are methods that have seen interest from industry to address other specific use cases.

The interest shown by industry in the companies and technologies above (among several others) signal growing awareness of the maturation of energy harvesting as a feasible and differentiating product. In the market developments above, a common model is for partnerships or consolidation through acquisitions of energy harvesting companies with industry semiconductor or equipment manufacturers, with the aim of making these components simply another part of their product line. Because of the lower power consumption of devices and the resulting gap for energy harvesting to fill, companies active in this space are seeing more interest in commercial deployments of their solutions, which will consequently help in the mainstreaming of their products.

What these market developments show additionally is the growth of energy harvesting beyond primarily consumer applications. Energy harvesting applications have already seen considerable adoption in wireless switches, with leading companies including EnOcean or ZF, the latter in partnership with ON Semiconductor to commercialize the solution. The reason for prevalence in this particular use case is its simplicity: the energy generated from the kinetic motion of the switch is enough to cover the data and power cost of a transmission on a low-power connectivity protocol, addressing the fundamental reason why energy harvesting has not seen widespread adoption so far, namely the mismatch between the energy that can be generated and the power requirement to do meaningful IoT applications. Once additional sensors, MCU processing, and more frequent measurements are required, the mismatch frequently becomes too large. The examples above show clear industry and commercial intent to be well-positioned with regard to these technologies as they begin to show feasibility, both from a technological and business standpoint.

It will be a long time until we see energy harvesting sensors deployed at massive scale in IoT applications. There are multiple reasons for this. First, the sources of energy are frequently unreliable: effectiveness of harvesting varies considerably based on, for instance, the levels of vibration or the frequency of radio signals harvested, which cannot be guaranteed or will not remain the same all the time. Second, the technology only works with the lowest-power components; companies such as ON Semiconductor are at the forefront of bringing the lowest-power modules to market, but until these are more widespread, commercial uptake of energy harvesting solutions will remain specific to certain ultra-low-power applications. Third, the technology works only with the lowest-power connectivity protocols, such as in BLE or in some LPWAN use cases. Everactive has even developed a proprietary protocol, with all others too high-power; this enables impressive capabilities, such as continuous reporting at 15-second intervals, but limits the solution in terms of mass adoption. Fourth, batteries are cheap and reliable, and enable lifespans beyond 10 years in simple use cases, which will be the primary target segment for energy harvesting solutions.

While the technology has the potential to be transformational, the market calls for real-life solutions. If a solution works and makes business sense, as in the case of simple switches, the market will adopt it; several other technologies have a way to go to prove this business case and get the solution to scale. Energy harvesting suppliers should take a use case-by-use case approach to address specific scenarios, rather than hoping to differentiate themselves on technology alone. Several of the market developments noted above take this approach, with Perpetuum a prime example of a vertically focused company with a product addressing a specific business case. Condition-Based Monitoring (CBM) more broadly is a field where energy harvesting technologies have the most potential for uptake, given regular vibration on many kinds of industrial machinery and equipment. Without this type of design-constrained approach, suppliers will struggle to find where they can add real value over existing solutions.