Does my smart home hurt or help the planet?

So Stacey on IOT recently produced an environmentally-focused blog which really made me think about how I use my home automation as a force for good. Over the last few years I appear to have accumulated quite a range of smart technology in my home now: routers, cameras, zigbee hubs, light sensors, PIRs, mesh routers. According to Imperial College London, turning electrical equipment off makes a big difference, but can using smart electrical equipment with some sensible automation achieve even more? In short, does my smart technology help the planet or hurt it? 

Photo by Nicolas Jehly on Unsplash

Am I right to be using newer devices like Zigbee or Thread instead of WiFi?

Clearly there are different types of device. Routers and sensors just sort of sit there burning carbon, and it’s difficult to see how emissions from these can be reduced aside from using solar/renewable electricity to power them in the first place. But, I am slowly replacing higher energy WiFi products with lower energy Zigbee/Thread products as items fail. I figure the energy saving from the newer devices almost certainly beats their production overhead.  

But for me the most relevant question is are Zigbee products + a gateway lower energy consuming than WiFi products (ignoring the WiFi router which is surely a “fixed cost” of being online)?

A report into the subject from GE via GreenTech Media concluded “...the ZigBee chipset/module consumed about 0.39 watts over the 24-hour period, compared to 0.87 watts for Wi-Fi” so Zigbee data is roughly half the energy cost of WiFi data. But I do need a ZigBee gateway which is a WiFi device so practically, if I have six Zigbee devices per router then here’s the rough calculation:

  • Zigbee plus gateway = 6 x 0.39 + 1 x 0.87 = 3.21 W per Day
  • WiFi = 6 x 0.87 = 5.22 W per Day
3.21 vs 5.22 W per Day

So my Zigbee setup theoretically uses about 61% of the equivalent WiFi setup and, clearly depending on the product. And I can choose to use rechargeable batteries instead of a power supply which I would do wherever possible.

So most of my home energy use is pulled via light bulbs, heating (we don’t need cooling in the UK) and cooking. I will deal with heating and cooking in another post but for the purposes of this discussion I will focus on the detail of lightbulbs.


I know I always talk about light bulbs but which ones should I be using?

The current generation of LED light bulbs promise significantly less energy overhead (and hence reduced carbon emissions) when compared with their halogen/incandescent predecessors and also appear to offer a long life of service. And so it seems entirely logical, to me at least, to replace high consumption halogen bulbs with low consumption, long-life LED alternatives, even if there is some embedded carbon, rare earth elements and non-recyclable materials utilised in the production of the new LED bulbs. However, given that halogen bulbs last hardly any time at all this still makes sense.

And there are a few options out there: Hue, LIFX, Nanoleaf and a whole universe of options on the Tuya platform which works natively with Apilio, as does almost any Zigbee 3.0 bulb. 

I originally chose LIFX as they were talking about a 22-year life span, which seemed almost too good to be true (and it definitely was for at least one of their “outdoor” bulbs) and the light brightness is amazing and no hub is required. But they are rather pricey, the app is shockingly bad (on Android at least) and they are annoying to own: upgrades take too long, frequently fail and frankly I have found them more trouble than they are worth. 

But more and more now I find myself wondering about how much additional power those LED bulbs are *really* taking over their lifetime. Are they as low energy as they claim? Fortunately, this blog on light bulbs by Future Proof helped answer the question - they seem to have done a thorough job on researching the carbon emissions from a number of bulbs and worked out Hue and Philips had the lowest emissions over their respective lifetimes. I do have to say I would like to see the new Zigbee/Thread-controlled Nanoleaf bulbs compared alongside as, in theory at least, their standby power consumption should be significantly less than the WiFi bulbs (which is one of the main reasons the Hue bulbs beat the LIFX bulbs below in the table of annual CO2 emissions per bulb).

And most importantly, how does my automation help?

I also have to ask myself what exactly am I doing with all these devices. How am I using them differently and are they actually saving me energy or do I just feel better thinking they are?

Here are some typical use cases (in my daily life) I found to help me determine if I’m being green or (Mr) Bean. 


Six zigbee LED bulbs in a room, controlled by PIR and a light level sensor. If there’s movement in the room the lights stay on else they go out after 3 mins of inactivity. Previously, as it got dark - let’s say 18:30 - people would turn lights on and leave them on all night until about 22:30.

  • Automated version energy consumption:

6 times a night x 3/60 hours x 6 bulbs@11W = 19.8WH

plus 0.87 WH hours for the for the PIR sensor, the light sensor and the Zigbee gateway for 24 hours = 21.5WH 

  • Previous version energy consumption:

4 hours x 6 bulbs@11W = 264WH

  • Previous version energy consumption (incandescent bulb):

4 x 6bulbs@60W = 1,440WH

21.5 vs 264 vs 1,440 WH

Wow! That’s a 12x saving over lights on all night and a 67x saving over old incandescent bulbs. So I reckon I am *definitely* doing good for the environment with my lighting automation - which doesn’t even include when the kids would occasionally leave lights on ALL night every now and again. Now the lights won’t let them ; )

As an aside, and per my argument above, I am going to ignore the embedded carbon in the light bulb as the old bulbs needed changing so frequently anyhow. 

And reviewing this actually made me think, do I really need a light sensor in each room when actually, one sensor somewhere useful probably tells me if I need the lights on everywhere. Kerching! A few excess light sensors gone. So actually this exercise will really make me think more about how to minimise the number of sensors required in future!


External lights only turning on when people show up. Previously, lights go on at sundown and off at 22:30.

This is virtually the same as the previous example except the lights and the PIR sensor are WiFi as they’re too far away from the nearest Zigbee gateway. Also there is no light sensor as it’s a sundown (+20mins offset) to sunrise (-20mins offset) Apilio routine - gosh I LOVE our time offsets.

  • Automated version energy consumption:

4 times a night x 3/60 hours x 6 bulbs@11W = 13.2WH plus 0.87 WH hours WiFi PIR sensor for 24 hours = 14.1WH 

  • Previous version energy consumption:

4 hours x 6 bulbs@11W = 264WH

  • Previous version energy consumption (incandescent bulb):

4 x 6bulbs@60W = 1,440WH

14.1 vs 264 vs 1,440 WH

Wow! That’s a 19x saving over lights on all night and a 110x saving over old incandescent bulbs. So I reckon I am definitely doing good for the environment with my lighting automation.

But what about server consumption I hear you ask? What’s the overhead of using the Apilio web app? Well we are pleased to estimate that each of our users is likely to be consuming less than 5 WH of energy per month on the Heroku servers, making this energy usage almost entirely negligible when compared with our appliances at home. And if you think about how many users we have, making the sorts of energy savings I discuss above, then I think we can conclude that the combination of IoT devices and Apilio are doing their bit for the planet!

By Dan @ Apilio

April 28, 2021
Projects and Thoughts on IoT

Join a fantastic IoT community

Head over to our Community to join the conversation, get help with your setup and show off your build to inspire others.

Join the Community →

Make your life smarter today