Honestly - a Pi of some type. They have tons of IO options; serial, I2S, SPIO, USB, gpio, dedicated radio/lorawan plugins, etc.

If there something you want to interface it with, odds are that someone has built an interface hat to do it.

You can get power relays, sensors, the works.

It's a huge ecosystem, and because they are used a lot in education the documentation is usually pretty good.

Also big bonus of being solid state, small and relatively low power.

I've done a few bits and pieces with them - a home power monitor (interfacing with both WiFi power sockets as well as 433mhz broadcast devices using the rtl-sdr dongles), remote control of power sockets in my office (by being a 433 broadcast devices itself for said power control sockets) as well as a real-time data display for a legacy, serial-only, automotive ECU.

Good idea - I've never gotten into PIs much...

OTOH, I have several board sets for the NewBridge data switches I designed years ago. A 6809 CPU board with ROM and 48k of battery backed up RAM (and I've modified my MON09 debugger to be able to auto-start code loaded into the RAM). And 8-port serial cards (which can effectively be 48 I/O lines - not many wires to hook the cards together and "instant" fairly powerful SBC!

And the 09 is kinda ideal - I have my own Micro-C compiler for it (essentially the same one I have written tons of DOS stuff with - including the X10 "CM11" interface code).

Although just tossing together a instant-boot (from GoTek "floppy") PC looks promising too... Would let me use the code I've already written with almost no change, and could have an 8x25 text display I could see via the camera.
- Not too worried about power - not much beside mainboard and would be powered ON only when I'm accessing it.

Still trying to decide the best/easiest way.

Don't forget if you intend to use wake function you need a "green PC" capable motherboard, usually later ones with the 3.3V and an SL enhanced CPU, then it should just wake on LAN with the cable hooked up, when you access it... though you do need a hub/router/switch capable of sending wake up packets.

Anyway, all sorts of ways you can go, like look into Wiznet ethernet serial bridges with really dumbass pocket change uC boards. Look at old touch tone decoder control devices and rig a modem to work them, or an audio channel over net.

Then there's a lot of android apps that make things easy and controller hubs that glom together functionality of multiple smart plugs etc. There was a site on an .io domain which was basically click to build an app for remote control of an Arduino to do what you want.

Probably too many ways to do it, like a telepresence bot that you just ram into pushbuttons or something daft.

I'm running Home Assistant-core running in docker on my NAS with an APC SmartUPS 1000VA upgraded with a pair of 18AH batteries to cover for power outages and initiate shutdown (as well as allow monitoring) through USB. With a fresh pair of batteries it can keep the system alive for somewhere between 4-6 hours.

HA has a lot of integrations for IOTs plus community integrations through HACS. The core/unsupervised version don't support addons officially, but there is a HA OS you can run from a faster thin client like the Futro S900 or the HP T630. The Fujitsu has more room inside if you want to add some hard drives to function as an NVR, though I don't know how well the HA OS plays when it comes to adding more functions outside of HA itself. It doesn't matter if you run it in a containter.

I'm not using cameras, but if I would, I'd consider Zoneminder.

Now that I've closed DDS I no longer need/want a 24/7 server, and I only need it (remotely) while travelling (3-4 times a year) ... I also don't want to pay-for/maintain a static IP - so I would need it connecting though a 3rd party server, which means "keep alive" messages (something AnyDesk does - and also what my Camera base does)... Which kinda prevent effective use of "wake on LAN".
-- But.. turns out I was mistaken and AnyDesk still supports Win7 and even further back (so it will likely continue to do so for some time).

So I've revised my plan. All the cameras/IRled will do now it start the server. I'll still use a small microcontroller so that I can count pulses to do "other things" if needed, prevent accidental activations, provide some LEDs of feedback, and shut-down on inactivity ... but I won't need it complex enough to run the X10 "CM11" or interface-to/control much else beyond digital "bits".

Fortunately I had already modified the server to have remote "digital" access to it's power-button, and tell if it's powered-on (something my homebuilt UPS used to perform "automatic gracefull shutdown during extended power-fail").

Once running, I can get full "console" access through my phone or tablet, which gives me back the ability to remotely control LOTs of local stuff in my home.
-- This includes a lot of "old school" stuff I've built over the years which has simple ie:not-modern (LAN, USB etc.) interfaces: RS-232 via my-own control protocols, X10, digital bits etc... Things that modern "off the shelf" remote setups don't much know how to interface to - and I really don't want to redesign all these (and make them more complex/unreliable)...

It will also let me view the local (non-IP) cameras that I have - really feels like the best solution (for me)
-- Still more than happy to correspond with anyone else who might also want this ability.

I've more or less have been doing stuff like this for a long time. I really like not having to depend on anyone else (being in the cloud and all the ills) to do anything as pretty much everything can really be run at home for free there are a couple of caveats that make it hard to integrate a system with modern conveniences that I know of now:

1) Email. If you want to send email without going through google/microsoft, at the very least you need a static IP. If you don't need this, that really uncomplicates alot of trouble you have to go through to communicate that way. Arguably, if you only ever send email to yourself, maybe then you just need dynamic DNS but that type of situation is not what I am mentioning here, I'm saying real send/receive with the outside world (even just receive) is a really nice to have if you want to be independent at home. So really, static IP.
2) Push notifications. If you are doing monitoring, which is what it sounds like, sending notifications to a smart phone can be really handy when you are mobile. If you never leave your house, you won't need it. But on the road this is a good option. You also need a static IP, realistically, even if it is only going to you.

Both of the above you can surrender to google/microsoft, or not need it, even better. Then you don't need a static IP.

Everything else, like remoting in, I'd use a dynamic DNS service. I think that part you understand.

So when it comes to just talking, certainly from the average device today, it would be good to have a small 24/7 server that is a linux machine, like a raspberry pi. I'm not a fan of raspberry pi, for cost / form factor, but they can be still very useful because it can speak all the modern protocols, giving lots of options for access, and be always there to enter from to manage the other components of your system alot easier. Perhaps you can suspend it automatically, and wake up when you send some kind of magic packet, but I don't see why one would need to suspend something this small. From there, you manage your other systems. And that 24/7 tiny system can do the "keep alive" and update the dynamic DNS service.

the3dfxdude wrote on 2026-01-29, 16:32:

I've more or less have been doing stuff like this for a long time. really like not having to depend on anyone else (being in the cloud ...

Me too, been doing remote access/control since the days of my first systems in the 1970s ... initially Modems and BBS related stuff...

I do want as much "full" control as possible.. In recent years, I've only been "dependant" on AnyDesk which gives you fill access to a WinBlows system. Seems pretty stable and consistant (I've been using it for many years). It nice to get "local" as that lets me access things on my home net without putting them on internet. Eg: I have several IP cameras which all want to "phone home" for remote access - I've blocked them from accessing inet in my router but can still view them with local browser.

I didn't mind leaving a system running (alone) while I was away - I live in a rural area which sometimes has extended power-outages ... so I built my own UPS which could gracefully shutdown the server, and wait for power to stabalize before powering it back up (I don't keep much else on my server and Win7 seems pretty good about powering up/down in response to the front-panel "power" button (which I've brought out to a connector on the back of my server)

Now that I've retired, trips can be longer (just got back from being away for a month) - would rather not rely on a PC running alone that long...

Which is what lead to this thread... I have a couple connected cameras which I can use to "push buutons" ... this does lead to an additional dependance on an external company, but I expect connected cameras will be available "forever" and can change it out if I need to.

I could also "push reset" and force power it off if things go bad...
- There is still of course the whole "unexpected crash - tell me by GUI that it's ok to startup" problem - "modern" systems really aren't designed with reliable remote access in mind!

I'm now thinging of modifying my UPS to NOT auto-power on the server, but accept another "button input" to start it, while still doing a gracefull power-down during power-fail. That way I can turn it on and use it whenever and wherever I want, and just shut it down as "normal" afterward.

1) Email. If you want to send email without going through google/microsoft, oat the very least you need a static IP.

In the old DOS days, I had a command-line email client that I could use... But now that it's all gone GUI - I played with providing serial input from a "mouse replacement" to auto "do stuff" on Winblows, but this OS is too inconsistant to make this really workable...

2) Push notifications. If you are doing monitoring, which is what it sounds like, sending notifications to a smart phone can be really handy ...

An interesting idea ... Know any good/simple way to send a text or notification under your control (by computer). ... I can set up the connecterd camera to send a notification if it detects motion, so I might be able to build something to "fake" it ...

I am in early stages of messing around with this https://tinycammonitor.com/ and the minimum it can do is send you motion alerts. Seems quite powerful. Gotta let the cams speak to the internet to get that, but you can still block their "phone home" addresses. However you can also set them up on just a local android box, but then you gotta figure how to pass on alerts. Also has a reddit group https://www.reddit.com/r/tinycam/

DaveDDS wrote on 2026-01-31, 12:37:

Push notifications. If you are doing monitoring, which is what it sounds like, sending notifications to a smart phone can be really handy ...

An interesting idea ... Know any good/simple way to send a text or notification under your control (by computer). ... I can set up the connecterd camera to send a notification if it detects motion, so I might be able to build something to "fake" it ...

Yes, so OpenHAB and HomeAssistant are services you can run locally that can monitor your cameras and you can create custom event triggers to send notifications. If you want to send SMS messages, you'll likely need a paid for service (again a third party). And if you use the phone app for OpenHAB or HomeAssistant to sent the event that way, the instructions you may find for push notifications might be for google. So if you do want to be independent of them, you need some kind of messaging server, which then you will need some kind of 24/7 computer running and preferably a static IP for anything sensitive enough the address cannot change. There are cameras that integrate with the OpenHAB and HomeAssistant, so the only things you need to provide are the devices and the messaging service you want to use.

I haven't tried auto-suspending my computer and having it wake up on an event. I feel like if it even did work it would introduce even larger delays, and small delays may be critical sometimes. So that is why I am saying, at least one computer managing everything 24/7 and a static IP somewhere for the best reliability.

Your use of a UPS is good for the case of a power outage. Again I am not sure if triggering a boot/wakeup from another device will be too slow or not, so I'd rather go 24/7 and always power on when power is applied for the managing computer.

DaveDDS wrote on 2026-01-31, 12:37:

In the old DOS days, I had a command-line email client that I could use... But now that it's all gone GUI - I played with providing serial input from a "mouse replacement" to auto "do stuff" on Winblows, but this OS is too inconsistant to make this really workable...

Interesting. I have done serial/modems in DOS plenty but nothing too complicated. More recent days I tried doing serial stuff modern Windows, and while it works, are you saying it is not really workable? I haven't considered Windows as I really never need to use it. OpenHAB and HomeAssistant provides scripting to automate things and so that might be a better way to go than trying to fake stuff through GUI interaction. And if you are on linux maybe the GUI will no longer be in the way.

My stuff is also a work in progress and I'm trying to figure out some solid options similar to you. If you want to kick around some ideas, maybe we can share.

the3dfxdude wrote on 2026-01-31, 16:33:

...Your use of a UPS is good for the case of a power outage. Again I am not sure if triggering a boot/wakeup from another device will be too slow ...

I keep the systemfairly "lean" on "stuff that loads at boot" - so it's typically <30 secs, and would be no worse remotely than when I boot it up it in my basement...
I'd like faster, but it's quite reasonable. And I only do this remotely from "time to time" ...

Interesting. I have done serial/modems in DOS plenty but nothing too complicated. More recent days I tried doing serial stuff modern Windows, and while it works, are you saying it is not really workable? I haven't considered Windows as I really never need to use it. OpenHAB and HomeAssistant provides scripting to automate things and so that might be a better way to go than trying to fake stuff through GUI interaction. And if you are on linux maybe the GUI will no longer be in the way.

Sorry - I mispoke with "serial" (although I've "faked" serial mice inputs lots - very easy)
Nearly as easy, and much more acceptable to more recent systems is "PS/2"

The pic shown below is a gadget I built years ago to automate certain functions (very handy in some games)

It generates PS/2 keyboard and mouse signals (with the shown PS/2<->USB adapter I could use it via USB as well), has a big keypad on the front (with a little display so I could both configure it and prompt for "next in sequence" in certain cases),. The audio jacks mean it could "listen" for audio feedback. The "ext" inputs support varios ON/OFF external sensors (which could be as simple as a very large RED button on the desk). There is also an I/R receiver in the upper-right front so I could have it listen for remote codes)

It can do "most anything" you could do manually (and then some), but the big problem is it can't "see" the screen, and other than audio clicks and beeps, it's "flying blind".

Keyboard input is simple, and mouse is not as bad as you might think... It keeps track of where exactly it has placed the mouse pointer, and for cases where Winblows might have moved it, you can reposition it to position 0,0 by moving the "mouse" far upper-left.

Worked really well to simplify "doing complex things" when you were there actually using the system, but Winblows tendency to "pop up" unexpected prompts unrelated to what you are actually doing made it work less well for fully automated remote operation.

I just though to go look for it - wasn't sure I still had it. It has an STM32 inside it, and I have no idea where the code would be now... I may see if I can resurrect it.

BitWrangler wrote on 2026-01-31, 15:16:

I am in early stages of messing around with this https://tinycammonitor.com/ ...

I started with a number of different "IP" cameras... D-Link, VstarCam, Tenvis etc... They all wanted to maintain a connection to a different companies server...
I blocked them all at the router, and could still view them from a browser locally. Using "AnyDesk" I could effectively "sit" at that local computer from anywhere in the world (using another PC, a tablet or even just a phone **)... This was quite a few years (decade) ago - so all these are considered out-of-date by most --- but it all still works fine!

We've "modernized" with "Eufy" cameras - these can be configured to connect directly to an external server like the others, but you can also get a "base" which all the cameras connect to, and has it's own hard-drive so recorded video isn't saved "in the cloud" - all the Eufy server does it facilitate connections from their app and that base (no static IP required - but it does maintain one
"full time" connection).

**I have a Microsoft "universal mobile keyboard" which can bluetooth connect to three different systems, and a Logitech "touch mouse" which is tiny and can bluetooth connect to two different devices -- so my tablet or phone "feels" like using a PC (ok, screen might be a tad smaller than usual on the phone - but high resolution makes it very workable)

DaveDDS wrote on 2026-01-31, 21:47:

... wasn't sure I still had it. It has an STM32 inside it, and I have no idea where the code would be now... I may see if I can resurrect it.

Well... I found the code... (was a bit tricky, it was buried in a virtual machine) - looks like I last worked on it in 2017 (yeah, I know this sounds not terribly long ago ... but it was before my major injury/memory-loss... so it's a bit of a chore to recover exactly what I did!)

If anyone is interested, I can (eventually)* share what I have:

-

It can operate a computer system fairly capably via keyboard and mouse input, and can "listen" for audio output and read other input signals (digital or analog).

It uses my own ESL (Extensible Script Language) which has a "compiler" to generate a binary "executable" from source code. This is a fairly capable script system with variables(numeric and string), as well as many other statements, including all the usual "program control" type constructs (if/else, while, do, switch, call/return etc...)

This device makes it fairly easy to automate computer operations which were not designed to be automated. EG: Require manual operation via a gui.

-

*"eventually" because as noted above I'm having trouble recalling exactly what I created (funny thing - my more recent memory was much more affected - if I had built this 20 years ago, I'd probably remember exactly what I had done)

... and it was obviously a "quick and dirty" project, I had documented my ESL language itself, but I have no notes of how exactly this thing funcitons or what the various script extensions do exactly (other than their names) - and no real comments in the extensions codes.

So... I'm going to be spending a few weeks "reverse engineering" and documenting my own work to some extent... I'm kinda keen to do that anyway even if no-one else is, I see lots of use cases for it!

Things like that, sometimes it just comes back when you get "hands on", trying to think it through in advance gets you nothing, start messing with it, bingo, all falls into place.

Absolutely ... but I do have "recent year" memory issues, so we'll see how it goes!

And just to help others undrstand more about what I'm talking about... here's some more pics:

Inside - it obviously was a "quick and dirty" creation - you can see the STM32 board along the top - The keypad was one I salvaged from a old dead desk calculator... no encoder, just a wire matrixm scanned by the STM32 - you could just use some "push buttons".

I also found the serial cable I used (to load scripts into STM32 data flash) and even my "big RED button" (yeah, it's just a big bottle cap with a board mount button inside - I kept this beside my mouse/keyboard .. for some games it was nice to be able to very quickly trigger a "recovery" operation just by "pounding you fist on the desk"!)

Ah.. some of the memories are starting to come back (at least for this)