6 ms·

Designing a Wireless Device That Lives Forever

59 points by adunk 9y ago

retSava 9y ago
Hi! I'm Marcus at Thingsquare, excited to see this post hit HN :) Feel free to ask me anything, or hit me up at marcus@thingsquare.com if that suits you better. Best, Marcus
SlowBro 9y ago
Useful article! There's one thing you didn't consider: Charging temperature. Lithium charging temps are typically 0 degrees Centigrade. Any lower could damage the cells. NiCd is a bit better and there are some low temp NiMh but for the ultimate go with lead acid. They are not small but there are cells the size of a D battery.
- retSava 9y ago
  Thanks! You are correct, the li-poly I've looked at specify basically the same operating temperatures, discharge: -20 to +60 deg C charge: 0 to +40 deg C Exactly, the energy density is much less with lead acid, and they also have poor low-temperature discharge performance. I wonder at what level a very insulated and slightly heated (if temp < 0) li-poly would be better? I see there is still room for improvement in the system model we've used in the article :) Stay tuned for the sequel! We plan on doing one on indoor-solar powered devices too :)
  - SlowBro 9y ago
    Since charge time occurs during the middle of the day when both ambient temperatures are higher, and sunlight is warming the enclosure, and since charging slightly warms cells, and since electronics also warm cells, you have a better chance of reaching the minimum low charge temperature. Maybe add some aerogel insulation? Also, active heating could work; requires a larger panel though. This solar-powered Pi project integrates a battery heater: https://hackaday.io/project/13260-lifepo4weredsolar1 https://hackaday.io/project/13260-lifepo4weredsolar1
SlowBro 9y ago
Also consider not doing mesh if you need a longer life. I would expect that LoRa modules could cover the same distance a mesh could, and would only transmit then power down. Some receiving would be done as well but not as much as a mesh. By contrast a mesh node is listening and transmitting quite often.
- retSava 9y ago
  Yes, a transmit-only approach would be the most power-saving approach here, but you'd lose the ability to adjust configuration on-the-fly, and firmware updates over the air (FOTA), among other things. LoRa has excellent long-range properties thanks to sub-GHz, modulation, and low bit-rate, but you can just barely send something back to it and certainly not FOTA it. The device used in the example post (and here: http://www.thingsquare.com/blog/articles/sensortag-power/ http://www.thingsquare.com/blog/articles/sensortag-power/ ) doesn't route even if part of a mesh. Ie, it's powered down for the most part so won't route packets for others, but it can still be many hops away from the gateway.
  - SlowBro 9y ago
    Sorry I didn't mean transmit ONLY, just primarily. Of course you can still receive updates with such a device.
  - SlowBro 9y ago
    Things Network describes a method of updating LoRa device firmwares: https://www.thethingsnetwork.org/article/firmware-updates-over-low-power-wide-area-networks https://www.thethingsnetwork.org/article/firmware-updates-ov...
luma 9y ago
Another issue with solar charging in colder climates is snow accumulation on the panel. Combined with less overall incident sunlight due to shorter days and cloud cover, charging issues with lithium batteries at temps below 0C, and the fact that most common air quality sensors of interest (PM) are comparatively power hungry, it makes fielding a "Lives Forever" device here in Michigan something of a challenge. The traditional approach here is large panels with large lead-acid batteries but the size of the devices involved bring with them further challenges with siting and maintenance.
- retSava 9y ago
  And dust/sand in eg desert-close environments too!
jcun4128 9y ago
Do you need a gateway to access the internet. I understand the problem of FCC, don't want to use the wrong frequency, but it seems even if your device can live forever it depends on an existinng network. When I think of wireless communication in the wild I would have to buy a GSM module and subscribe to at least an SMS service. Is there such a thing as free communication that isn't regulated. Maybe that's what Walkie Talkies are but assuming "internet like purpose" or at least remote in/out. That would be cool though just connect to it assuming your device could interface with it (spread spectrum?) I think I saw something about phones being able to make their own local networks as they're able to transmit like up to a mile. It was I think related to apple phones. Edit: to try to clarify what I mean by gateway, to connect to the internet need an ISP, not talking about the TCP protocol but literally have to "be allowed" to communicate. I know sounding like a conspiracy person I don't know how to explain it. Also I understand towers transmot far, intercontinental cable,satellites not saying to replace that.
- retSava 9y ago
  You need a gateway more or less since the devices don't use the same physical means of communication as the other ones on the network. The mesh network devices in this post use a 802.15.4-compatible radio, which doesn't work with wi-fi on a physical layer, so the gateway translates this (and IPv4 <-> IPv6). All RF communication is regulated (broadly speaking). On the relevant freq. bands here, the ISM bands on 868/915 and 2.4 GHz, the bands are unlicensed (you don't need a license) but you have to comply with certain rules. You must not send too strong, or too much, or occupy too wide freq. band. You must be able to cope with other transmitters. Re phones and connections, I think what you mean is ad hoc mesh networks using the wi-fi that the phones (and laptops) have built in. Then you get a network in which the devices can communicate with each other (share files, messages,etc), where they otherwise might have required internet to reach eachother. It's a different (but interesting) thing...
  - jcun4128 9y ago
    Thanks a lot for the information. I had a relevant experience with regard to video piloting you are supposed to get an Amateur Ham license and then not transmit (supposedly) more than 25mW but many of the transmitters are close to 1W if not more and depends on antenna design like helical. Anyway lots of information here thanks.
TeMPOraL 9y ago
For values of forever equal to ~5-10 years? I'm not criticizing the article, which is pretty solid, but covers only the power harvesting topic. I'm just commenting on the title ("Designing a Wireless Device That Lives Forever"). I would very much like to see a serious discussion about making electronics live decades. The primary concern would be internal degradation and environmental damages. How to make batteries that last 100 years? How to make solar panels that last so long? How to deal with the fact that today's electronic components are purposefully not designed to last long? That after many years, your uC will likely just give up and die?
- adunk 9y ago
  Yes, this is an interesting topic! And even if you can make the hardware survive for a long time, even in harsh conditions (something like this comes to mind: https://sploid.gizmodo.com/this-old-ass-commodore-64-is-still-being-used-to-run-an-1787196319 https://sploid.gizmodo.com/this-old-ass-commodore-64-is-stil... ), the surrounding software ecosystem will still be an issue. Everything from compilers for building the firmware to the server side software will have progressed/aged in those decades.
  - HeyLaughingBoy 9y ago
    Good Engineering Practice suggests that you archive the entire toolchain in case you need to make updates 15 years later.
- HeyLaughingBoy 9y ago
  today's electronic components are purposefully not designed to last long Well, how long do you want them to last? These things aren't designed in a vacuum, you know. If the customer is willing to pay for a CPU that will run reliably for 50 years, someone will produce it. However, given that most consumer electronics are considered disposable, don't be surprised that the components that go into them aren't designed to last very long. Engineers -- good engineers, that is -- design to Requirements. If the requirement is that something last for 15 years and the Reliability engineers can predict lifespan to +/- 10% accurately, then a good manager will say "make sure it lasts at least 18 years." That ensures a 10% margin beyond what Reliability predicts and it makes sure you don't waste time building a gold-plated product that costs more than people are willing to pay.

59 points by adunk 9y ago

Hi! I'm Marcus at Thingsquare, excited to see this post hit HN :) Feel free to ask me anything, or hit me up at marcus@thingsquare.com if that suits you better. Best, Marcus

SlowBro 9y ago

Useful article! There's one thing you didn't consider: Charging temperature. Lithium charging temps are typically 0 degrees Centigrade. Any lower could damage the cells. NiCd is a bit better and there are some low temp NiMh but for the ultimate go with lead acid. They are not small but there are cells the size of a D battery.

retSava 9y ago
Thanks! You are correct, the li-poly I've looked at specify basically the same operating temperatures, discharge: -20 to +60 deg C charge: 0 to +40 deg C Exactly, the energy density is much less with lead acid, and they also have poor low-temperature discharge performance. I wonder at what level a very insulated and slightly heated (if temp < 0) li-poly would be better? I see there is still room for improvement in the system model we've used in the article :) Stay tuned for the sequel! We plan on doing one on indoor-solar powered devices too :)
- SlowBro 9y ago
  Since charge time occurs during the middle of the day when both ambient temperatures are higher, and sunlight is warming the enclosure, and since charging slightly warms cells, and since electronics also warm cells, you have a better chance of reaching the minimum low charge temperature. Maybe add some aerogel insulation? Also, active heating could work; requires a larger panel though. This solar-powered Pi project integrates a battery heater: https://hackaday.io/project/13260-lifepo4weredsolar1 https://hackaday.io/project/13260-lifepo4weredsolar1

SlowBro 9y ago

Also consider not doing mesh if you need a longer life. I would expect that LoRa modules could cover the same distance a mesh could, and would only transmit then power down. Some receiving would be done as well but not as much as a mesh. By contrast a mesh node is listening and transmitting quite often.

retSava 9y ago
Yes, a transmit-only approach would be the most power-saving approach here, but you'd lose the ability to adjust configuration on-the-fly, and firmware updates over the air (FOTA), among other things. LoRa has excellent long-range properties thanks to sub-GHz, modulation, and low bit-rate, but you can just barely send something back to it and certainly not FOTA it. The device used in the example post (and here: http://www.thingsquare.com/blog/articles/sensortag-power/ http://www.thingsquare.com/blog/articles/sensortag-power/ ) doesn't route even if part of a mesh. Ie, it's powered down for the most part so won't route packets for others, but it can still be many hops away from the gateway.
- SlowBro 9y ago
  Sorry I didn't mean transmit ONLY, just primarily. Of course you can still receive updates with such a device.
- SlowBro 9y ago
  Things Network describes a method of updating LoRa device firmwares: https://www.thethingsnetwork.org/article/firmware-updates-over-low-power-wide-area-networks https://www.thethingsnetwork.org/article/firmware-updates-ov...

luma 9y ago

Another issue with solar charging in colder climates is snow accumulation on the panel. Combined with less overall incident sunlight due to shorter days and cloud cover, charging issues with lithium batteries at temps below 0C, and the fact that most common air quality sensors of interest (PM) are comparatively power hungry, it makes fielding a "Lives Forever" device here in Michigan something of a challenge. The traditional approach here is large panels with large lead-acid batteries but the size of the devices involved bring with them further challenges with siting and maintenance.

retSava 9y ago
And dust/sand in eg desert-close environments too!

jcun4128 9y ago

Do you need a gateway to access the internet. I understand the problem of FCC, don't want to use the wrong frequency, but it seems even if your device can live forever it depends on an existinng network. When I think of wireless communication in the wild I would have to buy a GSM module and subscribe to at least an SMS service. Is there such a thing as free communication that isn't regulated. Maybe that's what Walkie Talkies are but assuming "internet like purpose" or at least remote in/out. That would be cool though just connect to it assuming your device could interface with it (spread spectrum?) I think I saw something about phones being able to make their own local networks as they're able to transmit like up to a mile. It was I think related to apple phones. Edit: to try to clarify what I mean by gateway, to connect to the internet need an ISP, not talking about the TCP protocol but literally have to "be allowed" to communicate. I know sounding like a conspiracy person I don't know how to explain it. Also I understand towers transmot far, intercontinental cable,satellites not saying to replace that.

retSava 9y ago
You need a gateway more or less since the devices don't use the same physical means of communication as the other ones on the network. The mesh network devices in this post use a 802.15.4-compatible radio, which doesn't work with wi-fi on a physical layer, so the gateway translates this (and IPv4 <-> IPv6). All RF communication is regulated (broadly speaking). On the relevant freq. bands here, the ISM bands on 868/915 and 2.4 GHz, the bands are unlicensed (you don't need a license) but you have to comply with certain rules. You must not send too strong, or too much, or occupy too wide freq. band. You must be able to cope with other transmitters. Re phones and connections, I think what you mean is ad hoc mesh networks using the wi-fi that the phones (and laptops) have built in. Then you get a network in which the devices can communicate with each other (share files, messages,etc), where they otherwise might have required internet to reach eachother. It's a different (but interesting) thing...
- jcun4128 9y ago
  Thanks a lot for the information. I had a relevant experience with regard to video piloting you are supposed to get an Amateur Ham license and then not transmit (supposedly) more than 25mW but many of the transmitters are close to 1W if not more and depends on antenna design like helical. Anyway lots of information here thanks.

TeMPOraL 9y ago

For values of forever equal to ~5-10 years? I'm not criticizing the article, which is pretty solid, but covers only the power harvesting topic. I'm just commenting on the title ("Designing a Wireless Device That Lives Forever"). I would very much like to see a serious discussion about making electronics live decades. The primary concern would be internal degradation and environmental damages. How to make batteries that last 100 years? How to make solar panels that last so long? How to deal with the fact that today's electronic components are purposefully not designed to last long? That after many years, your uC will likely just give up and die?

adunk 9y ago
Yes, this is an interesting topic! And even if you can make the hardware survive for a long time, even in harsh conditions (something like this comes to mind: https://sploid.gizmodo.com/this-old-ass-commodore-64-is-still-being-used-to-run-an-1787196319 https://sploid.gizmodo.com/this-old-ass-commodore-64-is-stil... ), the surrounding software ecosystem will still be an issue. Everything from compilers for building the firmware to the server side software will have progressed/aged in those decades.
- HeyLaughingBoy 9y ago
  Good Engineering Practice suggests that you archive the entire toolchain in case you need to make updates 15 years later.
HeyLaughingBoy 9y ago
today's electronic components are purposefully not designed to last long Well, how long do you want them to last? These things aren't designed in a vacuum, you know. If the customer is willing to pay for a CPU that will run reliably for 50 years, someone will produce it. However, given that most consumer electronics are considered disposable, don't be surprised that the components that go into them aren't designed to last very long. Engineers -- good engineers, that is -- design to Requirements. If the requirement is that something last for 15 years and the Reliability engineers can predict lifespan to +/- 10% accurately, then a good manager will say "make sure it lasts at least 18 years." That ensures a 10% margin beyond what Reliability predicts and it makes sure you don't waste time building a gold-plated product that costs more than people are willing to pay.