I am a software developer new to IoT & hardware, so apologies if these questions are basic. I’ve browsed existing threads and have not found what I’m looking for yet, so let me know if I missed one.
I am setting out to build a small child GPS tracker as my first learning project, and I want to make sure I make an informed choice on hardware.
Here are my goals/requirements for the device:
Cellular enabled, works in Italy (assuming Hologram SIM’s got me covered there)
GPS enabled (it is a kid tracker, after all).
Sends location updates to cloud every 30 seconds
Has a battery life of at least 10 hours.
Device is as small as possible, a child could wear on a lanyard, maybe even watch or attached to clothing.
Having watched this video, and looked at this tutorial, I think I need an assembly with the following:
microcontroller
cellular radio
GPS breakout
battery
antenna
… and eventually a housing to wrap it all up.
I plan to program this assembly with a simple python script that reports data to the webserver every short time interval. Nothing fancy.
Some assemblies that I think may work based on the links above:
Given that my use case is in Italy, is it problematic or beneficial to be using a 2G cellular connection as opposed to more recent LTE-M / NBIoT? The Pycom solutions seems to be only 2G … does that mean I’m buying obfuscated hardware, or on the otherhand, how do I know which networks in Italy even support the newer LTE-M stuff?
What are the power considerations of each assembly … which would have the best battery life?
As I am not a hardware guy, I would strongly prefer integrated solutions for both power efficiency and minimizing failure points. Are there any solutions that integrate all three requirements (Microcontroller + GPS + Cell)?
As a DIY project, how small can I reasonably hope to get this thing? Can I get it down to the size of a deck of cards? the size of a watch? a key fob? I know commercial products get small, but I don’t know how much custom board design they do for it. For example, would love to know how they built the lightbug tiny gps tracker.
I appreciate answer to one or any of the questions above, thank you and sorry for the long post!
Hi Peter,
I am new to this too, but I have done something similar.
In general terms what you want to do is defiantly possible.
By that I mean there are products on the market that have the functionality you are after.
However it may not be feasible given your budget and requirements.
Eg the apple watch does those things and more. you can by one for a few 100 euro but if your were to make one from scratch…
2G is going away slowly, I am not familiar with its status in Italy. It also is heavy on power which would be my reason for not using it.
The Pycom FiPy supports CAT M1 and NBIoT it also support LoRa, Sigfox which maybe more appropriate for your project given the small amount of data you are transfering.
It is my understanding that the dash is dead. We spent a fair amount of time and resources developing a trial based on it. We went to buy 50 for a broader trial and discovered they were discontinued…
We moved to the pycom boards mostly for power consumption. The deepsleep is reasonably good (the dash was not particularly good on power)
I have no experience with the Adafruit Fona 808
If it were my project I would go with FiPy and PyTrack (the other pycom dev boards are cheeper but lock you into a communication protocol where the FiPy gives you the flexibility to experiment with all of them. )
I don’t use GPS much. My projects are monitoring equipment that is generally fixed to the ground but disconnected from power for several months before it is moved. GPS would be nice to have but it draws to much power.
DIY size is limited by convenience taking the Pycom example. The FiPy+Pytrack+aerials and a battery. should comfortably be less than 2 decks of cards.
You have a bunch of extra features and connectors that you don’t need but you didnt have to make your own board. If you wanted you could buy an OEM module from pycom and the components you need from the pytrack, design a board and go from there maybe get the size down to a match box or slightly bigger depending on the battery. (30 second cloud updates and gps reads is going to drain the battery, it may still be bigger than a deck of cards)
Thanks for the great response! Totally agree with what you have to say, but also wanted to add my two cents. However before I do just wanted to say that the Dash is back thanks to our friends at POEM technology.
@pejowei, blease see my answers to your questions below:
Given that my use case is in Italy, is it problematic or beneficial to be using a 2G cellular connection as opposed to more recent LTE-M / NBIoT? The Pycom solutions seems to be only 2G … does that mean I’m buying obfuscated hardware, or on the otherhand, how do I know which networks in Italy even support the newer LTE-M stuff? In italy we ofer 2G with Vodafone Omnitel and WINd, but like Dane points out 2G is an old protocol taht is on it's way out and is also very power hungry. Ideally you'd be using Cat-M1 or NB-IoT. However, at the moment we are still working out roaming arrangements for those networks in Italy and hope to have them done by early 2019. Cat-M1 and NB-IoT are protocols designed for low data low power usage so pretty much exactly what you need.
What are the power considerations of each assembly … which would have the best battery life? To optimize for battery life you will definitely want an MCU and modem with some sort of sleep option. All three options offer this so I don’t think you can go wrong with any. As far as the specifics go you’d need to look at the documentation for each and do a little math.
I haven’t seen a three in one device that is an MCU + GPS + Cell Modem, buy Cell Modems + GPS are quite common. An example of that is the Botletics SIM 7000 shield.
As I am not a hardware guy, I would strongly prefer integrated solutions for both power efficiency and minimizing failure points. Are there any solutions that integrate all three requirements (Microcontroller + GPS + Cell)?
As a DIY project, how small can I reasonably hope to get this thing? Can I get it down to the size of a deck of cards? the size of a watch? a key fob? I know commercial products get small, but I don’t know how much custom board design they do for it. For example, would love to know how they built the lightbug tiny gps tracker.
If you want to get to the design your own hardware level you can definitely get this to the size of a matchbox. Using off the shelf programmable hardware I’d say a deck of cards is reasonable.