Nerves can run on tiny devices and I’m thinking of ways to use it - and I wondered how would YOU use it if you had the chance/time?
My current thinking is some sort of smart camera that is activated by movement or sound - ideal for security or even wildlife observation. Could even be used for mega long time-lapse photography perhaps
I also love @Qqwy’s idea of a photo-booth! A friend and I were addicted to these last time we went on holiday and it’s a shame our local clubs and pubs don’t have them around here - they are a good laugh
Knowing nothing about how medical devices qre usually coded, I wonder if Elixir would be reliable enough for medical devices such as pacemakers, prosthetics with engines, insulin pumps, deep brain simulation, etc.
Elixir is not capable of real time performance, of course (scheduling is implicit and you can’t predict how many instructions will execute per second, etc). Actually, for most medical devices which don’t (or shouldn’t) accept a network connection, I don’t see how useful elixir would be.
But it would be fun.
Real-time medical device programming
I think it would cool to control an LED array to show system stats like build status on a bunch of projects or represent memory usage on a host.
I have the hardware sitting next to me for monitoring CI/CD statuses, it’s next on my list I was going to have it done next month but my nerves talk was accepted for ElixirConf, so that’s where my time is now going.
A post was split to a new topic: Are there any special considerations to keep in mind when buying hardware for a Nerves project?
Any kind of home automation! Like monitoring and collecting a database of power consumption, automatically turn lights on and off, maybe manually accept a wireless command to arm the security system when going out, maybe even plant irrigation.
The BEAM being rock-solid should be a selling point in such scenarios.
(And yes, maybe medical devices.)
I’d love to make a zwave controller
which would also play well with this
I just got my solar system setup! 4-300watt mono-crystaline panels and 4 150ah deep cycle batteries.
The controller unit comes with an RJ54 port labeled RS232 and an rj45 to rs232 cable… quite usles as i dont have any laptop with an rs232 port… and i wont want to stand in front of the controller each time just to view readings…
I have Raspberry-Pi 3.0 (or 2.0 not too sure)
I would like to build a setup where readings can be collected continuously and uploaded to my online SQL Server from where I can have a proper real-time display of power consumption charts on the go, as well as alerts when the mains power goes off or on.
I’m working (very slowly, a couple of days per year) on a game called DuckTag. It’s basically laser tag but with RFID tags instead of lasers, you’d have a tag reader on your back that you are trying to protect and an RFID tag on your hand that you try to tag others with. When tagged, you’ll have to go to a respawn point to get back into the game.
Currently it’s at the point where the RFID reading works, the devices boot and network with each other, and start an Mnesia database that handles all the data updates. Very rudimentary game logic has been added, mainly to test things. There’s issues with reliability though, it feels like sometimes the devices aren’t running their supervisors or get stuck on something and I have to boot them up many times.
Next step would probably be to modify it so that I can just start all the nodes on my laptop in terminal screens instead of the actual hardware, to make it easier to develop. It’s too much work to set up all the boards and chargers and memory cards every time (since it’s 5 boards minimum: 2 players, 2 respawn points, 1 “judge” that coordinates things).
The pretty horrible code is available here: http://gitlab.com/nicd/duck-tag
It would actually be awesome to program in something nerves-like for my propeller board, it has 8 (though 1 dedicated to handle message passing) cores of a pure message passing arch with no shared memory, but it is significantly not powerful enough for something like nerves, so I use my own homegrown coroutine C++ setup. ^.^;
However, it’s great for controlling lights and automation and such. I can dedicate a core or so to specific polling purposes with ease.
I want a USB/Bluetooth bridge for my wired keyboard. Can’t find a good commercial solution. Rpi0W ought to work great - small and cheap.
This is something i want Nerves to be better at. I built a device similar to what you are describing, but backwards, and Nerves doesn’t exactly support the pieces you need for it to work out of the box.
I wanted to use the Xbox One controller on my original xbox. (don’t ask why) turns out the Xbox one controller uses bluetooth, and the original Xbox controller is usb 1.1, so i just stuck an rpi0w in between them in gadget mode and every thing pretty much fell into place. Problem is, Nerves by default doesn’t expose configurable usb gadgets, or bluetooth so i had to do some work getting that up and running.
Me too! I have a main Fibaro HC2 based system at home which I’m mostly satisfied with, but there are a number of devices that are not well (or at all) supported by it. For this reason I also setup a RaZberry hub on a RPI3.
I have on my someday list a project to look into how much effort it would take to talk with the razberry device directly from Nerves. In theory it seems to be doable and would open up a lot of possibilities.
For your power-monitoring project, you’d probably just need to use a multimeter to figure out which wires on the DB9 serial port are connected to which RJ45 wires, then make your own cable to connect it to the Raspberry Pi UART pins. From there, you’d need to figure out what kind of protocol it uses, either from the manual or from “sniffing” the traffic on the wires while using any software they provide with the equipment. Reversing the protocol would certainly be the hardest part, if they don’t provide you any documentation or clues. The hardware part wouldn’t be too complicated.
Thanks, I will try to do just that to figure out which are the connected data cables on the RJ45.
Also i’ll try to get a USB to RS232 cable to read the data on my PC first, to figure what data the power controller is pushing out.
Just thought of something that might be a good fit for Nerves - food manufacturing!
Some of my friends have been telling me to start selling my homemade liver pate as it’s quite different to what’s out therein the shops - and I’ve been experimenting with make larger batches while also thinking about how I could do it at scale. I have deduced I would need an intelligent system because I don’t think strict quantity/process/time would be enough!
Am I on to something? Could this be a nice niche for Nerves?
Well it already helps with food growing, so i don’t see why manufacturing would be much harder.
FarmBot is awesome Connor and yeah, looks like some of those sensors could definitely help in mass food production as well! Wonder if the FarmBot team have looked into it? It could be quite a lucrative market?