I've been lurking on these forums for a while and doing a whole bunch of research for a project that's been on my mind for the past few months: I'm trying to put together a device that would ideally :
track current GPS location, without the need for storage (see next pt)
be able to communicate over GPRS (or 3G, but my understanding is that 3G boards are still a little bigger, and 2G would do the trick perfectly for very small data packets)
all that in the smallest possible size
Main goal is to affix that stuff on my cat's collar and have my own web server parse timestamped GPS data to map everything... hence the size constraint.
I'm not familiar with PCB design and will likely never be (and I've ready a bunch of threads mentioning that, if chinese stuff do the trick, it'll be cheaper and sizeable... but that was in 2013-2015). I am however very familiar with C/C++ development and as such I'm looking for a board design that would already implement what I need. The software part, I'll handle, as well as 3D-printing an enclosure.
So far I've come across two 32u4 boards that seem to do nearly what I'm looking for:
Arduino feather FONA: size would be perfect to horizontally attach to a collar, but SIM800H does not support GPS. And I don't see any equivalent with a SIM808.
SIM808 FONA boards : while these lookg great for functionality, they will require a microcontroller. Two boards to put on top of one another, that becomes bulky...
Elecrow 32u4 and A7 chip: seems to have been used for similar projects (see e.g. this github) but square/rectange-shape feels a little more problematic for a cat
The newest version of that same Elecrow stuff with an A9 chip yields the same dimension issue
I'm sure most people here are much more familiar with current hardware: is there an alternative that would retain GPS in a shape that's more like the Feather FONA ? I'm obviously eager to hear any other suggestion as well.
You are not going to be able to build anything small and light enough to fit onto a cat with an Arduino.
As well as GPS and GPRS you need a power supply and the whole thing needs to be rugged and waterproof.
Waterproofing (rather, water-resistance-ing) will be fairly easy to handle. I have easy access to 3D-printing devices and will just design an enclosing with a short space to place an o-ring around the edges.
Your point about small+light is however valid and the reason why I discarded solutions based on a arduino+daughter board. Power supply: I was thinking of a 1S LiPo element, anything above 1000 mAh should do the trick for, say, a full day of operation, maybe two. That would maybe go up to a weight of 25-30g, max ?
Hadn't thought of that one, I was under the impression battery that SIM80x implemented AT+CBC so voltage could be read. But indeed, wouldn't want to risk setting the cat on fire.
Funny idea, see where the cat has its meatings.
I have built a unit using an UNO and a NEO6M GPS for speedmeasuring. I notice that the data given are less good when more or less dence forrest surrounds me. The GPS works very well the more open fields there are.
I guess the same quality aspect applies to the position data.
GSM positioning works by querying nearby towers and triangulating from there (most likely using signal strength and other indicators to balance within the polygon?). From past experiences with A-GPS positioning (which I'm pretty sure is what's being implemented in all these chipsets), GSM is used as a way to restrict the area where GPS fix should be acquired, by downloading relevant almanac and ephemerides. Getting a gps-only fix, on a moving creature, with a GPS antenna that would be a few cm-long is a very, very, very long shot.
I'm guessing what you have on your UNO involves a more decent GPS antenna
That definitely is an option. Reverse engineering these Chinese stuffs can get quite cumbersome due to lack of documentation though.
However that's exactly why I posted here : to get suggestions for possibly other hardwares that would be related to this topic. Thanks for the heads up!
Not sure where you're from and if the same in your country but over here a lot of providers are starting to slowly drop the 2g network. Maybe something to keep in mind.
I have a cheap tracking device with a lot of functions. Sending a sms when movement is detected. Sending sms when given area boundaries are being passed. Possibility to call the device and listen to the surroundings.
Downside is when active tracking it drains the battery fast. Not sure how big the battery is.
If I remember correctly, 2G networks in France are still scheduled to be maintained for the next 4-7 years. That issue has crossed my mind, but this seems to be much more than the life cycle of a hacked or handmade device here.
3g alternatives will make the "size" challenge even more complicated (from what I've seen so far).
It has a somewhat an adequate form factor, includes battery (safety assessment will ensue when I get the thing in my mailbox), and based on decent chipsets. I am a bit concerned about the software part (I don't think I have the skills to properly reverse engineer it so that I completely cut ties with their crappy tracking website). Initial plans are to change the domain it's sending packets to, receive them on a custom server, and see from there if I'm able to build some custom UI that would just query Gmaps API and overlay data points.
OK so... a little longer than expected, but chinese stuff takes a while to arrive and I've just got the package yesterday. The form factor is definitely OK for a cat.
I tested the stuff quickly yesterday in the evening and did not manage to get a GSM signal and at first I feared it might be because my carrier (which i plan on using with this because they have that really cheap 2€/mth plan with 2h voice + 50 Mo data, which is enough for sending GPS locations from time to time) doesn't have thei own 2G network but instead do roaming on Orange FR. And I as afraid this chinese stuff would have a poor modem implementation and fail at roaming.
But this morning, it seems to be getting a signal! Now onto looking at the extended documentation. This thing gets configured by texting it. My first take will be to set up a web server of my own and analyze the packets it's sending. Once I get that running, I'll try to keep it documented here and on my Github!
I went with two slightly different models (figured I'd may as well order two considering how cheap they are). A ZX612 and ZX303, which apparently run of the GT06 communication protocol for GPS devices. I haven't tested yet the battery life etc, I'm really at the early tests of running a Python server on my laptop and trying to make sense of what is being sent
OK, so, it's been a while, but I've now had enough time to make sense of most of the packets that are exchanged, and get a prototype working on my cat!
The code is available on my GitHub, for anyone interested. I know this is not really a "using Arduino" project, but since I was asked here to keep people updated... There it is!
Thanks for updating! I'll have a look at your github but wanted to respond first ).
Like I asked before, how is battery life doing? Are you live tracking your cat? Or only when it's moving using the onboard gyroscope? Or only taking when a "GPS fence" is breached?
Maybe that's information to be found on your github so will head over there now first. Hopefully I can do something with my tracker too.
@Bringamosa: indeed, a lot of question I didn't properly address. I haven't tracked the cat in a real uncontrolled setup for now except an hour or so. I did however run the tracker in my coat for a few days, it lasted more than 4 days using the embedded gyro to enter sleep mode when not moved around (ie when I was at work, most of the day). Over a weekend, it did last for say a two full days + a night out (involving a decent alcohol consumption and walking around).
I haven't worked on implementing a GPS fence for now as the main goal is really to have a GPS track, for educational purposes! I did however download the sources of an open-source project called Traccar, enabling such features, but with somewhat limited support of these Chinese stuffs where communication protocol isn't really standardized.
@KASSIMSAMJI: no, I hadn't. Which ones did you have on mind, specifically?
