A common question when considering energy harvesting solutions to wireless powering of devices is: "...why can't we just use batteries? Surely with batteries lasting longer these days, being energy dense and being cheap these will do nicely."

Batteries, however, have three major limitations:

- Reliability
- Environmental impact
- Lifetime and cost of changing

Reliability
There are a number of areas where the use of batteries to power wireless devices, often sensors of one sort or another, is impractical because batteries do not provide a high enough level of reliability. A typical cause of this will be high environmental temperatures. In industrial sensing situations there are therefore applications where and energy harvesting solution provides the missing reliability. These will tend to be high value but often niche applications.

Environmental
The widespread use of batteries has created many environmental concerns, such as toxic metal pollution. Battery manufacture consumes resources and often involves hazardous chemicals. Used batteries also contribute to electronic waste. Any developments that lead to the proliferation of batteries is therefore likely to lead to enormous environmental concern.

Lifetime
Whilst many batteries are rated for a typically 3 year and sometimes up to 10 years lifetime, the reality is that in most real world applications they last significantly less, often months rather than years. Lifetime is dependent upon both the duty cycle of the tasks to be powered and on the environmental conditions under which this happens. Even if never taken out of the original package, primary batteries can lose 8 to 20 percent of their original charge every year at a temperature of about 20-30 degrees C by self discharge. Newer chemistry and modern lithium designs have reduced the self-discharge rate of rechargeable batteries to a relatively low level (but still poorer than for primary batteries). Start making the battery do some intermittent work and this lifetime reduces drastically making battery change a necessity in almost all applications. This is by far the biggest argument for energy harvesting.

This is an issue for those niche applications with wireless sensor nodes placed in hard-to-reach locations (e.g. inhospitable and remote or hazardous industrial environments or even medically implantable devices) making the changing of batteries regularly both costly and inconvenient.

However, it is also an issue on a massively larger scale for pretty mundane applications where the access to the batteries is perhaps not difficult due to remoteness but is nevertheless impractical or expensive. Where EH will come into its own is in powering all of those applications where battery change is just inconvenient due to a combination of cost and volume of such changes and the unpredictability of battery lifetime.

For example, EnOcean, a German energy harvesting company, already have energy harvesting powered solutions in >100,000 buildings. These include light switches, temperature controls, window contact sensors, lighting control - all elements of a smart building which enable them to produce an average annual energy savings of 60%. In just one example they have installed 4200 wireless and battery-less light switches, occupancy sensors and daylight sensors in a new building construction in Madrid. These are powered by energy harvesters and embedded in the building. This saved 40% of lighting energy costs by automatically controlling the lighting in the building, removed the need for 20 miles of cables, avoided the use and changing of 42,000 batteries (over 25 years) and saved 80% of the cost of retrofitting a wired solution.

Even more mundane is the example of US energy harvesting company that was approached by a company that runs the toilets in various public buildings e.g. airports. This company has a problem with the taps that use infrared sensors to detect hands and switch on water. Evidently they regularly stop working as they exhaust the 3 cell batteries that power them generally in around 3 months. Every time one of those taps goes down they get a call and have to send in a qualified plumber to scrabble under the sink and change the battery at a reported cost of $75 a time. There are undoubtedly countless other trivial examples that aren't about hazardous areas, deep sea, underground access, just inconvenient and expensive to change batteries due to who can or will change them. Another example would be the need that hotel chains have to periodically go around changing batteries on the key cars door locks on its rooms. There are current project investigating energy harvesting solutions based on piezoelectric materials and the pressure of inserting the card. Another simple but elegant solution to the cost of regularly having to go up and deal with door locks with dead batteries.

Texas instruments have developed EH test kits to complement EH suppliers devices and cite the applications in a wide range of areas including remote patient monitoring, efficient office energy control, surveillance and security, agricultural management, home automation, long range asset tracking, implantable sensors, structural monitoring and machinery/equipment monitoring.

So there are already millions of opportunities in currently battery-powered or direct wired devices where energy harvesting could save considerable sums of money in terms of labour. However, the even bigger opportunity is in the so-called 'Internet of Things', an ecosystem of wirelessly connected devices many of which have to be mobile and/or inaccessible for battery change. Cisco claim that already by 2010 the IoT contained 12.5bn connected devices, that by 2015 this will be 25bn and by 2020 it will be 50bn devices.

Cisco predict that as many as one trillion devices will become connected to the Internet. Such connections include sensors on electrical lines to conserve energy, RFID tags attached to shipping crates to track inventory, temperature sensors in lab refrigerators to preserve medical supplies, tiny humidity gauges scattered on a forest floor to assess fire danger, sensors drifting in the ocean to track pollution, or body worn sensors connecting your vital signs to a remote healthcare provider. For IoT to reach its full potential, sensors will need to be self-sustaining. Changing batteries in billions of devices deployed across the planet and even into space is not an option and furthermore the disposal implications of billions of waste batteries is huge. Energy harvesting technologies therefore need to step up with a solution that allows fit and forget powering of the device.