I have looked at LoRa/GPS combinations occasionally for the past couple years, since I have young boys that like to hike. Recently, the youngest got farther ahead than expected. We had FRS radios and told him to wait, but instead he tried retracing his steps and took the wrong path because he panicked. Fortunately, he passed within shouting distance.
So, I have started looking at LoRa/GPS again, but the prospect of tackling hardware and software is a bit daunting. I stumbled across Meshtastic by way of Andreas Spiessās YouTube channel and am happy to see that the T-Beam seems to handle the hardware side.
I write code (mostly Go, but a little Java and assembly too) and would be happy to help develop features that facilitate my particular application, but would like confirmation that the hardware would support it and that the community would welcome it. First, is my understanding correct that my kidsā T-Beams would not need a smartphone to send GPS coordinates? Second, could I write code (in the app and the firmware) that would allow me to send a special message (ping) that would prompt an automatic reply with my kidsā GPS coordinates? As a poor alternative, could I set up the firmware on my childrenās devices to send out their coordinates every 5 minutes? Finally, how difficult would it be add a hardware SOS button so that the kids can let me know they are in trouble?
Am I correct to think I could connect all my kids T-Beams to a single phone for initial setup, but only keep my T-Beam connected to the phone so that I can see them on a map?
Welcome! I will try and answer what I can after only a couple weeks of casual use.
The kids T-Beams do not need a smartphone as it has an onboard GPS antenna. However there wonāt be a way to send a message to you, but they can read yours on the device. And a SOS button sounds like a great idea!
The location frequency I believe is set to 15 minutes by default, but pressing a button forces a location check if Iām correct. Iāve set my frequency to the 30-60 second range with success, so 5 minutes should work great.
Unfortunately I canāt offer any knowledge on coding!
But you are correct. I use a pair of T-Beams, mine is synced with my phone over Bluetooth to show messages and positions on a map, where my dogs is a standalone unit operating on the same default channel.
Iāve had an average of 1000m range in heavy trees and uneven terrain, and 9km line of sight from hilltop to hilltop. The nice thing is when you donāt have a signal, you still have a heading to their last known position.
I am copying a post I saw from @mc-hamster ! A Pelican 1010 Micro case fits perfectly, but I also made some rough ones out of PVC pipe. Thereās a thread on some DIY cases people have made. And some great 3D printed ones our there too.
Using the python interface you can change to GPS update frequency.
Everything you have mentioned has been discussed here before.
A lot of work lately seems to have focused on making a more generalized platform, and splitting things into different layers, and using plugins for specific functions.
From what I remember, the GPS updates havenāt been optimized, and donāt get sent often due to the air time requirements.
Perhaps you could look into working with @geeksville and @mc-hamster to create and use a base Meshtastic location API that could be used by different plugins to request, or provide a location.
I think once locations can be provided from a plugin it will be much easier to start optimizing. There are a lot of tricks available to maintain accuracy but drastically reduces bits transmitted.
If you are not traveling across the globe you really donāt need as many bits if your nodes agree on a given region, then only provide updates in relation to that area.
Only sending the delta of the last position could save a lot of bandwidth, but some logic needs to be in place so new nodes can request the need information to use said deltas.
Another plugin could then be created for people that want to record tracks with more updates, and then a detailed track could be transmitted or revived upon request.
I figured my use case was likely to be discussed before. I was trying to quickly get a feel for the current capabilities and community direction before committing to purchasing a few T-Beams and installing the development tools. Now I just have to wait for shipment. . .
GPS update frequency is totally configurable from within the Python API. If you donāt use the slowest / longest range setting, you can get by with fairly frequent updates.
For just tracking a single device that donāt require the longest range the current location setup is going to work was mc-hamster said.
For people that want to track more than a device or two an overhaul of the location functions is needed. The current implementation doesnāt scale well.
Tracking of races and live stock are two frequently requested features that would be much more practical once location updates donāt require as much air time.
Is there back-of-envelope computation on how frequently we can update GPS coordinates, as a function of number of nodes and channel settings?
Iām mostly interested in the case of 5-10 nodes with slow/long settingsā¦ I would like to be able to send coordinates every 50-100 sec or soā¦ Is it reasonable?
Also, I am curious, is the current limitations due to the LoRa bandwidth limits, or per-device duty cycle imposed by regulations, or mesh protocol overhead? (I guess all of the above )
)