I was reading this article, "Five Building Blocks of Self-Powered Wireless Sensor Nodes" (shared on IoT Meta) about energy harvesting in IoT.

It lists a couple of harvestable energy sources for example:

  • thermal energy
  • vibration energy

I am planning to implement some prototype, harvesting devices to explore the possibilities. To get some experimental data about how much energy can be harvested in my home. Currently I am trying to identify the possible spots in my home where these energy harvesters could be placed and efficiently used.

What I though of:

  • Maybe a vibration energy harvester could be attached to the washing machine in the bathroom. Risky environment as the humidity can be quite high so appropriate protection is needed.
  • Thermal energy harvester could be placed near the oven in the kitchen.
  • Thermal energy harvesters could be placed near the heating elements in the bedrooms.

What other parts of the apartment should I place further harvesters for experimenting?

So for example can a microwave oven be a possible source or other frequently used kitchen appliances?

I have an electric boiler above my bathroom (inside the apartment), so maybe the hot water pipes running down to the bathroom are good sources as well.

What else could be there? Are there any more possibilities to harvest vibration energy in a home?

  • 3
    I noticed a few answers relate to equipment which has an existing mains power supply. Maybe these are less interesting, unless there is a reason that a self-contained, passive powered device is particularly valuable. Commented Feb 15, 2017 at 8:13
  • 2
    @SeanHoulihane Indeed, those options are less interesting as they require an other consuming device to be powered but still can be a good way to track household device states. Nevertheless I welcome all reasonable ideas. Commented Feb 15, 2017 at 8:22
  • 2
    @Goufalite Ghanima covered it pretty well. I did not mention light on purpose as my apartment is quite dark. Still such information is in scope and can be valuable for others or later for me. So you can go ahead if you would like. Commented Feb 15, 2017 at 13:24
  • 2
    @BenceKaulics one thing I haven't seen anyone mention is that for the quality of light you get indoors, amorphous solar cells are actually more efficient. Mono or poly-crystalline solar cells are much more efficient with direct sunlight, but their efficiency falls off much faster when light is indirect.
    – Aaron
    Commented Feb 15, 2017 at 18:13
  • 2
    @Aaron you're right, my answer does not cover it but the referenced paper supports your claim wrt a-Si cells.
    – Ghanima
    Commented Feb 16, 2017 at 20:41

16 Answers 16



I don't know if it is in the scope of your question but you can't imagine the wasted energy of opening/closing a door!

A French comic strip named Gaston Lagaffe written in 1970 introduced a person who harvested the energy of an opening door to press oranges, print a picture, make coffee,...

EDIT, thanks to to Olivier Dulac's comment I was able to find the strip without having to read all my Gaston collection! Translation if from me but I can't translate the last picture, it's a French pun on words about a unhinging ourselves when we are mad.

gaston comic strip

To stay in the scope of your question, I am thinking of some scenarios:

  • Footsteps? Harvesting vibration with a plaque where people walk the most : entrance, where you put your coats, near a Kinect playfield.
  • Computers, servers: harvest their heat.
  • Make a outdoor boiler using hot water roof panels (of course you need your landlord's agreement) and harvest the heat.
  • There are charging shoes!
  • 1
    Everything is in the scope and doors are just great, I should have thought of it and computers as well. Commented Feb 14, 2017 at 19:47
  • 3
    The footsteps idea works especially well in large public spaces; some electricity-generating paving slabs were used in a shopping centre in London, according to Scientific American.
    – Aurora0001
    Commented Feb 14, 2017 at 20:40
  • 2
    @Aurora0001 Can confirm; my dad was the managing consultant on that experiment, and it appears to have gone very well.
    – ArtOfCode
    Commented Feb 18, 2017 at 3:41
  • 2
    Great comic! I think you could roughly translate the last panel as "Coincidence: just as I went off my hinges, the door went off its own... Really!??" Not quite sure how to translate "m'enfin" though (I went with 'really' but Wikipedia suggests 'what the?' could work as well).
    – Aurora0001
    Commented Feb 18, 2017 at 10:51

Solar energy is even indoors an alternative to thermal and vibrational energy harvesting. This indoor photovoltaics use both the natural light and artificial light.

On the Feasibility of Indoor Light Energy Harvesting for Wireless Sensor Networks (Carlos Carvalho and Nuno Paulino) discusses indoor light energy harvesting of the attenuated natural illumination. The authors describe the powering of a low power wireless sensor node with an average power level of 11.51 µW. This low average power level is achieved by a very low duty-cycle of the system making one transmission every hour only.

Sizing Solar Energy Harvesters for Wireless Sensor Networks (Murata Application Note M1002) compares different ambient energy sources:


Clearly the use of ambient light in an indoor setting yields a lower return than other sources, still it could work out for certain use cases.


While having so much valuable answers to my question I have also done some further research about the topic.

I have found an additional harvesting solution and also a great reference design about a wireless, solar energy powered sensor node, without using a battery.

First the additional method from Texas Instruments, which is called RF switch harvesting technique. (The video is worth watching, the RF switch part starts at 2:29.) It is based on a magnetic switch which structure can be seen on the image below.

RF Switch harvesting model

On the drawing here in the red is the switch part the user will push. We have a magnet inside with the two poles. And this is creating a magnet field through this core back to the other pole.

When the user flips the switch, you can see the field changes. It goes the opposite direction through the core. And we know from Maxwell's equations that this change in magnetic field through these windings will produce a current. This is energy that we can harvest and make use of.

The switch produces an AC current that needs to be converted to DC current which can charge a capacitor. Then the energy stored in the capacitor can be used to power a low-power wireless MCU.

remote switch - power solution curves and flow

TI has actually realized this and they were able to supply enough energy for 21 milliseconds for the wireless MCU.

power harvesting diagram and visual setup

The reference design is a Solar Dice which can wirelessly transmit an accelerometer's data to a smartphone. The solution does not use a battery and quite compact. Worth reading or browsing the provided test data, schematics and other documents.


This is an idea I've had that I'll throw out here...

It seems like it should be possible to have a small device that you simply wrap a mains power cord around a few times (or wrap the device around the power cord, I'm not sure which). When the "host" device is powered on and current is flowing, the "parasite" on the power cord would become powered--not much, obviously. But, perhaps enough to periodically send a wireless "ping" with an ID (like a MAC address) to some receiver. What this would get you is a sensor that would tell you whenever "dumb" devices were switched on. Based on the frequency of the pings, you might even be able to infer how much power the dumb device was using.

Since the mechanism would be simple and inexpensive, and require no regulatory approval for plugging into mains power (UL/CE etc.) possibly someone could sell these in packs for a target price of ~$5 each. When you get enough of them (which would be the point to have a lot), a machine learning system could learn things about your habits: The hair dryer is on, it's probably ~30 minutes to bedtime. The coffee machine is on, turn on the radio...that sort of thing. And you don't have to buy a smart hair dryer or coffee machine that costs 30-40 times as much as a $5 sensor just to get that usage data.

  • 1
    excuse me for my ignorance.. will the parasitic use of the current's generated EM field actually make the transport losses bigger?
    – userfuser
    Commented Feb 16, 2017 at 11:53
  • 2
    If the power cable contains both live and neutral, the EM from one would cancel out that from the other, would it not? Hence no power to tap. Commented Feb 18, 2017 at 22:22

Radiators should be a source of vibration (driven from the pump), though you would probably need to investigate the resonance frequencies quite carefully.

It is also possible to embed kinetic energy harvesting in switches, although that may not be quite the use case you're imagining.

The other side of the equation is determining how much energy you require for a transmission, which will give you some information about how frequently it is possible to transmit, together with your leakage/standby drain.

  • 1
    Those guys for example have a piezo powered system for energy harvesting in switches enocean.com/en and they claim it to be useful for self-powered wireless sensor solutions for the Internet of Things.
    – Ghanima
    Commented Feb 14, 2017 at 20:32

This is a great idea to harvest energy from our daily energy use. But you have to consider that the energy you harvest comes from somewhere. If you want to harvest some heat from your room heating elements, you must be aware that all the heat converted to electricity via your device will not heat the room. So, your room heater will have to consume more electricity if you want to keep your room as comfy as it used to be before you set up your heat harvester.

  • Take heat from devices where heat is wasted (i.e. any object whose main role is not to heat, like computers of servers as said above) seems to be more in the spirit of what you want to achieve.
  • In a Gaston-Lagaffe-style, you can set up small devices on doors, stairs etc...to create energy every time you pass by. An idea among others, you can set up a small dynamo at the top of your stairs, and every time you go down you grab a nylon-cord and run the dynamo. Once downstair, you let go the cord and it rewinds. The nylon cord will run for about 20m if your stairs are 10m long, so with a good gear ratio and efficient dynamo you can generate the amount of power you need for any small device around.
  • 3
    I'm not sure about this issue of heat 'loss'. Provided the temperature gradient is within the room (i.e. not using a cold source from outdoors), and the electric consumption is within the room, there is no loss of heating. The good thing about electric heating is it all ends up as heat eventually. Commented Feb 16, 2017 at 18:15
  • 1
    @SeanHoulihane is completely right. Moreover, while heat off of computers is waste heat, moving it quickly off the computer is intentional and something you're expending energy (fans, pumps, peltier cooler, whatever) to do. Harvesting usable energy off of that amounts to providing resistance to heat transfer; as long as you're actively moving heat off (fans, etc.) then whatever energy you harvest is bounded by the energy being spent to move the heat off quickly, i.e. it's just a net waste. Commented Feb 16, 2017 at 22:10
  • 1
    I do agree with both of you. What I wanted for OP to notice is that there is no magic with energy creation, and what is created somewhere is necessarily taken from elsewhere, so take heat from a heater is not an idea as good as it seems. About heat from computers etc, as @R.. mentionned, there may be energy consumed to extract the heat. But on some servers, you can notice that the surface is hot, and so extract heat from the surface will not decrease fan efficiency or whatsoever.
    – le_daim
    Commented Feb 17, 2017 at 10:03
  • 1
    @DamienMarty: If the surface is hot and you put some obstacle (the harvester) between the surface and the outside air, it's going to heat up more, and the fans inside the computer case are going too work harder trying to get the heat to move off. Unless your cpu can run ok passively cooled even with a harvester on it, you are not harvesting the waste heat. You're harvesting the electricity that went into powering the fan, at a loss. Commented Feb 17, 2017 at 17:37
  1. WiFi signals

    Forget about stealing your neighbor's Wi-Fi to surf the internet. Using cheap everyday materials, researchers at Duke University’s Pratt School of Engineering have developed a remarkable device that can convert microwave signals, like those used to wirelessly transmit the internet, into usable electricity. So in the future, you might instead be able to use your neighbor's Wi-Fi network to power your home.

  2. EM Fields from AC wires behind wall (if mounted near an electrical wire).

    Abstract - There has been considerable interest in energy harvesting for wireless sensor networks. Energy harvesting from thermal sources such as body heat and mechanical sources such as human motion have been proposed. There are also sensor network systems that harvest energy from the visible part of the electromagnetic spectrum. However, ambient light levels in indoor environments are typically significantly lower than those found outdoors and highly dependent on the nature of the indoor environment considered. Recently, lowpower clock synchronization using electromagnetic energy radiating from AC power lines was proposed. In this paper, we go a step ahead and try to answer the question: Can energy be harvested from the electromagnetic energy radiating from AC power lines and use it to operate a wireless sensor network with a low duty-cycle? We find that such energy harvesting appears promising.

  • 3
    Harvesting WiFi or any RF signal is almost certainly a violation of FCC (or equivalent in your jurisdiction) regulations on producing harmful interference, since the harvesting is effectively interference equal to some negative scalar (or rather, likely, a different negative scalar per frequency component) times the signal being harvested. Commented Feb 16, 2017 at 22:14
  • 3
    @R.. I'm sorry but that statement is not correct in any way shape or form.
    – CramerTV
    Commented Feb 17, 2017 at 2:40
  • 2
    @R.. An 'energy harvester' is simply an antenna. You're suggesting putting up an antenna is somehow illegal or restricted. RF signals can be blocked in many ways, a vase, a couch, a pinball machine, etc. Placing items in your house that intentionally or unintentionally block RF signals is not restricted in any way. The device is not 'producing' interference. Devices must not emit 'illegal' radiation which a harvester would not. Further, the FCC actually requires that many devices MUST allow interference to be legal. electronics.stackexchange.com/a/35190/29607
    – CramerTV
    Commented Feb 17, 2017 at 3:19
  • 2
    @R.. Whatever energy your neighbours WiFi send into your home is lost anyway. You are not "stealing his energy", you are "picking up his trash and reusing it". Which is illegal someplaces btw.
    – Lenne
    Commented Feb 17, 2017 at 15:29
  • 2
    @R.. Putting an antenna on my spectrum analyzer, TV, radio, Raspberry Pi, etc. to capture radio waves coming into my house is not illegal or restricted. Nor is putting an antenna on my 'microwave to power' converter illegal. Good lord man. As noted in the paper, this is not limited to WiFi signals. You can capture TV signals, Radio signals, GPS signals - any RF that excites your antenna is potential energy to use - and not illegal.
    – CramerTV
    Commented Feb 17, 2017 at 17:56
  1. An exercise bike/ treadmill rigged to a dynamo. This will also give you an incentive to work out.
  2. Rather than using the microwave oven/convection oven which will eat into their function, you can place a small chimney over the cooking place (though the escaping vapors might pose a problem).
  3. I don't exactly know how, but people sit on the couch all day. I know that it's potential energy, but something could be achieved there.
  4. The door idea from the above answers is great.
  5. Tiny wind mills in the patio.
  6. Solar cells (available on the cheap now) can be used outside.
  7. Craziest: Using a small turbine under the shower/tap
  • 3
    The turbine under the shower is rather creative but care must be taken to prevent the sinker being clogged. The turbine might be an excellent hair collector. Commented Feb 16, 2017 at 18:10
  • 5
    Can you please elaborate the 2nd I am not exactly understand what is the point there. Commented Feb 16, 2017 at 18:12
  • 2
    Using the heat in an oven might increase the amount of power it needs to heat the food. On the other hand, a small chimney over the cooking stove would trap the heat that is escaping. [Chimney example] (faberindia.com)
    – SagarU
    Commented Feb 16, 2017 at 19:46

One often available temperature differential that might be harvested would be:

the difference between internal house temperature, and external.

With a temperature controlled living environment, there would often be a non-trivial differential, one way or the other. Even without that, external temperature often changes through 24 hrs, whereas we try to keep internal temperatures constant - So much of the day there still might be a differential. I know the differential is minor, - But the point is - It's there for a lot of the time.

A huge advantage of this, it would seem to me, for iOT sensors, is that the energy source is available at much of the periphery of the living space - Ceiling, windows, walls, floor (perhaps).

Then again, maybe I am only thinking this is sensible because I live in Australia and we have days above 40 degrees celsius. ;-)

I haven't done any calculations to figure out whether you could harvest enough energy... But obviously - Depends how much you need.

One upside is that being such a passive device, and planar in shape, you could probably design something fairly large, say, 6 inches square? but still aesthetically pleasing for ceiling or wall mounting.

One downside is that you may get days in a row where there isn't much differential - If the external temperature hovers around a comfortable room temperature.

Just a thought. Cheers.

ADDENDUM: Yes, I know this isn't free energy, but the OP didn't explicitly ask for that. A common problem with IoT devices is running wires everywhere. Perhaps the OP, at least partially, just wants to avoid that?

  • 2
    We had here some outdoor below -15 °C temperature here, that will also make some difference. Also the main purpose is not about getting free energy but eliminate continuous battery replacement, recharging, wires. I welcome all idea which widens the experiments. Commented Feb 17, 2017 at 14:31
  • 1
    I'd be interested if you could report back how practical this is. I have heard of IoT sensors lasting 2-3 years on a couple of AA batteries, so surely you could harvest enough to compete with that. But I haven't heard of the idea proposed much, so I posted!
    – spechter
    Commented Feb 18, 2017 at 12:07


The average human produces 100 watts of power at rest. That's enough to power about 6 laptops (assuming the estimate of 15 W for a low-end laptop) if you can convert 100% of the power you produce into electricity. Of course, we can't defeat the laws of physics, and inevitably you can't convert 100% of your energy into electricity; after all, you need energy for your metabolism and bodily functions.

There are startups creating felt that generates power from your body heat (approximately 1 or 2 watts can be generated from that).

Alternatively, you could use an exercise bike to generate electricity, as shown in this really nice infographic by the IEEE.

And, if all else fails, you can always hand-pedal your sensors:

Pedalling a radio

Source: Wikipedia (Public Domain)

This method relies on you being inside (and some methods rely on you actually being present to operate the crank/pedal/dynamo), so it's not entirely foolproof.

  • 2
    Oh my god, in the future evil machines will lock us in a virtual world and harvest our energy to power... DOOR SENSORS!!!
    – Goufalite
    Commented Feb 15, 2017 at 12:53
  • 2
    Unfortunately, no one can be told what the IoT is.
    – Ghanima
    Commented Feb 16, 2017 at 20:50

You mention generating electricity from heat. All devices that turn thermal energy into work have a hot side and a cool side. The theoretical maximum energy available from the heat comes from the difference between those temperatures. This means that cold things are just as good for generating energy as hot things are--but you'll do even better if you can get a hot thing very close to a cold thing and put a thermoelectric generator in between.

You also want to be careful not to harvest energy that you're paying to create. You could get some energy between hot water pipes and cold water pipes--but then you're just making your water heater work harder--might as well just use mains power and pay for it directly.

On the other hand, if you exploit the energy differential between a wastewater pipe (suppose somebody's taking a hot shower or draining pasta) and a cold pipe, then your cold water gets slightly warmer, and your wastewater gets slightly cooler--neither of these are likely to be a problem (unless the specific cold pipe you're using feeds into your refrigerator).

  • 1
    It depends on the cost equations. If there is a heated water supply, it may be overall cheaper to place a load on this system than install an electric circuit too (or replace the batteries assuming access is maybe tricky) Commented Feb 16, 2017 at 18:18

Snow sliding off the roof

This is a particularly northern answer, but if you have a metal roof with a fairly steep slope, and you get a decent amount of snow in the year, there is quite a bit of energy being wasted when it comes sliding off.

Alternatively, for people who get less snow in the year, the rainwater that comes off the roof is a good place to harvest energy. An electric dynamo could be hooked up to your gutters, producing energy any time it rains.

It could also be hooked to your Internet connection and monitor:

  1. The amount of rainfall
  2. The amount of energy generated
  3. A comparison between anticipated rainfall (from weather forecast) and actual rainfall

Potentially, you could even have sensors in the gutters to adjust the dynamo generation capacity based on the quantity of water flowing through.

  • 1
    Ha! An XKCD comic was missing in this question : what-if.xkcd.com/23 (middle of the page). "If the generator rig cost $100, residents of the rainiest place in the US—Ketchikan, Alaska—could potentially offset the cost in under a century."
    – Goufalite
    Commented Feb 18, 2017 at 9:04
  • 2
    @Goufalite If the point is just economic, you're right, and good point. However, many people harvest energy for other reasons, including independence from government controlled supplies and not wanting to waste available energy. Personally, I think your comment applies to the majority of the answers.
    – anonymous2
    Commented Feb 19, 2017 at 12:22

Your land line.

There is about 48V available on any land line in your house. Lots of bad articles on the internet about this, though this seems to be the best I've seen.

Part 68 of the US Federal Communications Commission's telecommunications regulations states that any device that connects to the phone line and is not actively communicating must present a resistance of at least 5 MΩ

...an image from the article listed above...

  • 3
    Not to mention that this might be unlawful depending on your legislation.
    – Ghanima
    Commented Feb 16, 2017 at 20:52
  • 1
    @Ghanima: Almost certainly, but probably half the ideas here are unlawful already... Commented Feb 16, 2017 at 22:16
  • 1
    @Ghanima: I'd have to wager most people here have done more unlawful things than use a couple microwatts of power from a telephone line. Any time you charge your phone in a public place you are stealing way more power than this.
    – Fuffer
    Commented Feb 18, 2017 at 3:23

Many kitchen and other household appliances have significant vibrations as well. Think dishwasher & microwave in the kitchen. Clothes washer & driers. Any bit of electronics with a motor (cooling fans? turntables?) likely has a notable vibration output.

  • 3
    All of these have mains power too, which makes them less interesting sources of low value energy. Commented Feb 15, 2017 at 8:12
  • 1
    Welcome to the site! Do you know if there's any data about how much energy could be harvested from the objects you suggest? I think that'd make your answer very useful for anyone with specific power constraints.
    – Aurora0001
    Commented Feb 15, 2017 at 9:46
  • 1
    @sean my interpretation of the question was that we were looking for a way for a parasitic sensor to draw energy without being connected to the mains. Which means anything that is connected, and wastes energy in the form of vibrations, heat, or light output, would be a potential power source.
    – Rozwel
    Commented Feb 15, 2017 at 17:49
  • 1
    @Aurora I have no idea what the specific numbers would be. This is not an area I am particularly familiar with, I just flipped over because it made the hot list, and added an answer because no one had mentioned this potential. However, from general observation, I would note that different models and ages of equipment will vary greatly in the level of vibrations they produce.
    – Rozwel
    Commented Feb 15, 2017 at 17:55
  • 1
    @Rozwel That's fine, just suggesting it might be a good improvement to your post if you get time to look the figures up. Good answer anyway.
    – Aurora0001
    Commented Feb 15, 2017 at 18:14

There are many options possible but in terms of easiest way to harvest my list would be:

  1. Thermal energy captured by thermoelectric Peltier elements. Thermoelectric elements are non-expensive and available in Europe, on Alibaba.COM and Amazon.COM.

    Please note that back in 1930 thermoelectric elements were used in radios in USSR in tundras harvesting energy from carbide lamps.

  2. Solar energy has nowadays very sophisticated and cool products. For example something like SolarWindow can be a motivator to use it.

  3. Piezoelectric (discovered in 1880 by Pierre Curie and his brother) polymers are really fascinating materials. For several years my team is playing with such material PolyPower introduced by Denmark company Danfoss.


A slightly obscure answer is 'from the sensor'. By making the hardware event driven you can potentially improve the standby power by such a large magnitude that a battery can be regarded as a non-replaceable part of your hardware.

See this research from the University of Bristol where they have optimised a switching element for pA leakage. Although I've seen this in passing before, I didn't appreciate the practical value till watching a video which described the scope for using the device.

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.