My wife and I started a meetup group for sporting dogs that need a lot of off-leash exercise - Vizslas, Weimaraner, GSPs, any pointer, and…any dog that loves running in a pack and has a nice owner. In 5 years it has grown to 12k+ members across the USA.
In our opinion a GPS tracking device is a must for any dog/pet - the cost is so low, the stress of a pet going AWOL is awful (for pet and owner) and can end in the worst, and completely avoidable, way.
The choice of good GPS trackers presents a challenge - real time tracking with terrible battery life, or great battery life with acceptable but not always reliable tracking that has a 2 minute delay.
I want to understand how GPS trackers work - but I have no expertise in electronics.
I wanted to start by taking apart existing GPS tracker products to reverse engineer and understand the tech.
If anyone has expertise/knowledge and would enjoy taking these things apart with me, let me know.
Honestly, thats probably not the right approach to begin with. I mean, yes, youll need to get into the hardware eventually, but the first thing you should do is learn about the mathematics involved. The actual technology is probably not going to be much more than a ready made chip, antenna, and power supply (plus some supporting electronics to connect to a cell phone network). There are youtube videos that can be very helpful in giving you a high level idea of how the tracking works.
The electronics themselves, while i most certainly wouldnt discourage you from exploring, are probably not really all that beginner friendly and i wouldnt be surprised if they were highly resistant to reverse engineering. But if you want to open one up and post high res pictures of the boards, ill be happy to tell you what all the parts are for, etc.
Hope this is helpful.
As an edit, maybe should also mention that im oretty sure there are premade modules to use with a prototyping controller, like an arduino or raspberry pi, and it might be quicker and easier to build a prototype from scratch than try to reverse engineer a commercial model. Thats not the case in every case, but i think it is in this one.
The tradeoff you’re describing is fundamental to any mobile communications system. A GPS receiver knows where on the earth it is. A GPS tracker combines a GPS receiver with a wireless communications mechanism of some kind, and radio transmission requires a discrete output of energy.
“Real-time” trackers are transmitting (ie consuming battery) more-or-less continuously. You can conserve your finite battery capacity transmitting with less energy (ie reduce your range) or by only transmitting from time to time (ie your 2-minute delay).
You could also use a different mechanism to figure out where you are, but GPS works basically anywhere and has benefit from decades of receiver optimization.
From some prior technical support of my dog’s FitBark, I got some information from the company directly that relates to some of the design trade offs they made in regards size, battery life, etc. I can dig up some of that info if helpful
For my dog I use an AirTag (I get them for cheap so I have lots of them). My thought is anywhere with cell reception that a GPS collar would be able to transmit its location from would likely have people with iPhones.
The communication needs to be long-range. There are a few options.
Most obvious, common solution is SMS network. It will not work in places with no cell service. You do have to buy a SIM card for it.
LORA modems another option. Long-range radio transceivers, can go more than a KM in the cities and much further in unobstructed fields. But it requires a matching receiver.
The other option is meshing to, like, any stranger’s open wifi when it gets near and then calls home. There’s no “open Bluetooth” where you could go looking for a stranger’s Bluetooth and use it to send a message across the internet. It would work with an app if someone voluntarily installed that app on their phone and left it running (like OpenPath), which does lower battery life.
Thanks Russell, much appreciated.
You’re right that I am not exactly up on the maths of it all and its a good idea to do that.
I think the main challenge is battery life vs power to track real time - and I’m sure the existing unit manufacturers (or marketing companies working with Chinese/Asian manufacturers) are also looking at this issue. I’m interested to find out if its a cost issue or if there is some technical barrier that no-one has cracked or is prepared to invest time to solve.
There are definitely modules available for an arduino or pi, but I am not sure that I am going to make anything better than already exists at the moment going that route. What I feel I (personally) need first of all is to understand what is different technically about the hunting type trackers that track in real time via GPS but suck battery life quickly and need a dedicated hand receiver vs the Whistle / Fi / FitBark type products that use GPS in conjunction with AT&T LTE-M networks, bluetooth when in proximity to the owners phone and wifi in the house and use an app on a phone.
I will happily take you up on your offer of sending you some hi res images when I take examples of the above apart.
Thanks Jon, and agree. The GPS receivers used on the hunting collars are highly accurate in real time. The issue I have seen is 1) owners can lose trust in the tracking collars because sometimes they are in weak LTE-M cellular areas and so end up returning the collar (which is really not a risk worth taking or a problem not worth tolerating for safety) or 2) they trust the receiver but it is very bulky, requires a dedicated handset and they forget to charge it every day so the dog/pet is rarely wearing it.
I know I am simplistically starting from a point of saying I want both accuracy and battery life. It is ultimately a size of unit, power source and owner controller issue to solve.
Thanks for the Hackaday link Matthew, I’ll be setting aside a good chunk of time to watch those. The AirTags (to me) are almost a no-brainer as to something people should use whether or not they have a tracker. This would be effective in urban areas which are very dangerous for a pet to be loose in because of traffic volume but have a lot of people forming the iphone network.
We’ve heard from many of our members however that they don’t trust them - a few people have tested the Airtags and have said they have placed one on a busy trail somewhere but had no success with anyone picking up the Airtag. A lot of this is about trust in the tech. If it doesn’t work very well people just won’t bother.
Interesting Danny, hadn’t thought about the SMS option. Could be a good back up to the main tracker possibly. Problem when a pet goes Awol is they mentally become feral often within a few hours - they normally go awol because they were frightened by something man made, and they associate any human with the fright so often take off into uninhabited or remote areas where by definition there is little to no cell service. however, they also often keep moving, so as soon as they touch even a very weak cell signal the collar will ping a location, as long as the collar still has battery life.
I don’t know much about LORA modems. I know the hunting GPS collars can have a range of up to 5-10 miles but that normally requires line of sight (and also to a dedicated receiver) so have limited reliability when the frightened pet goes into a rural, wooded and often hilly environment.
The issue is pretty much a fundamental tradeoff that I don’t think you’ll necessarily solve by taking apart trackers that are already out there. Instead I’d maybe try to think about reframing the problem/solution. For instance, I don’t know much about sporting dogs in general, but do they by chance tend to wear harnesses as opposed to just collars? If so, a harness with integrated GPS tracking would offer you significantly more space to integrate battery packs, a larger receiver or transmitter, etc. that could increase the usability without having to reinvent the wheel.
Thanks everyone for your response, I really appreciate it.
Takes a village etc.
A little more background than I felt I could put in my introductory message.
We have a lot of experience with 2 trackers in particular; the Whistle tracker and the Fi tracking collar. We (and many of our members) found Whistle to be unreliable, often fell off or just broke within 12 months or less and their customer service was almost non-existent. We have had a lot of success with Fi and continuously promote them to our members. Their battery life is category leading (can last 3 weeks or even in limited at-home only circumstances up to 2 months), their app is very good, the unit has had some design issues but they have been very responsive to us, though not always to our members.
The hunting collar we had (a Garmin) was excellent for tracking but so bulky and the battery life was hours only.
And hence I have arrived at a place where I can’t believe there isn’t a better mousetrap to be made but also respect the many people already involved for years in making one of the two main solutions and seeing none of them have been able to combine both benefits.
And this is why I started by thinking I would take one or two of each version apart to understand the tech. I can get technical specs/drawings etc. I am sure but there is nothing better than getting your hands on the things.
Anyway, thank you so much again for all your input. Feel free to throw anything else my way if you have any idea of any sort - all thoughts are welcome. I may well come down to Asmbly to take these things apart and will let you know when I will be doing that in case anyone is interested in joining me.
yeah, love this Matthew.
Again, trying not to drown everyone in super long messages but I am right with you in needing to reframe the problem.
So, sporting dogs when in the field will often wear an orange jacket of some sort, probably not a harness because it can get caught etc.
However, the issue with the dogs getting lost is that a tiny percentage of these breeds are actually used for hunting. Most owners have them as pets, often because the owners are active outdoors, hiking, long walks daily etc. and these breeds need a lot of off-leash exercise for both their physical and mental well being.
My strong view is that owners would happily have their dog wear something that is more than a collar if it was safe, didn’t hinder the dog or make it uncomfortable and significantly increased the likelihood of a lost dog/pet being found safely and quickly.
Another insight which sounds counter intuitive, is when a dog is lost one of the worst things the owner can do is go around shouting its name out - because the dog has almost certainly been frightened and taken flight - it associates humans with being present when the scary thing happened and it is usually the owner who is the closest human so the dog can very quickly become more wary of their owner than anyone else. Crazy, but true. The key is to know where the dog is and then figure out a plan to lure it to a place of safety which can include a trap.
Anyway, have definitely thought about how something bigger than a collar might help with, e.g. bigger power source, solar power, etc. No actual solutions yet but 100% agree with you on re-framing the problem - you are onto the point that looking at an old problem through old lenses sees nothing new
Just for reference, these are the breeds (Hungarian Vizslas, Weimaraner and German Shorthaired Pointers) we have the most of in our meetup group which is called Viz Whizz Dogs. The 2 dogs together are our two.
So i just want to make clear: i wasnt suggesting you build out a prototype just because. I have a strong suspicion youre not going to find any useful parts in the commercial trackers. The thing is that theyre probably just a specialized microcontroller, and all of the interesting bits will be hidden away in the chip or in software. In fact, its not out of the realm of possibility that the hardware in both coild be exactly the same, and the differences in performance and battery use could be solely due to software configuration. I was, instead, suggesting that building out your own prototype could give you something useful to experiment with to figure out why the draws are different, etc. I know for a fact that the cell phone baseband controllers tend to have proprietary software, and good luck with that. Broadcom (or qualcomm?) baseband controllers are notorious for that.
Im the last person to discourage taking something apart, i just have a strong suspicion it wont be at all useful to you.
