Thank you for posting this, it reaffirms what I've been thinking for so long. There are many opportunities to retrofit old systems of all types with modern low-cost embedded technologies.

Around 2019 or so I was approached by an engineer who had a small business retrofitting very old machine tools with modern motion controls. Think very large lathes and planers. The problem they had was that in order to get these systems working with newer controls, they had to make time-consuming modifications to the old machines, in some cases modifying the axis motors and that cost added up quickly. The engineer realized that it was theoretically possible to take the old analog position signals and convert them into something a modern motion controller could read. That converter box would make the retrofit pretty much plug & play, but he didn't have the programming expertise to make it happen.

We probably built the first iteration of the converter for under $50 in parts and less than 50 hours of development time. That had me searching for other similar opportunities. I have found a few, but they tend to be one-offs that aren't worth the time unless you're already building something similar. Either that or my ability to see opportunity sucks!

I think I see some parallels with the aftermarket for old cars. Holley's entire catalog is "here's a thing that'll modernize your 1960s engine without replacing the whole damn engine"

This kind of thing is a low-volume, low-margin business by nature. It's those kinds of businesses that the megalomaniacs of our field tend to ignore. There are people with needs, and money they are willing to exchange for a solution to those needs. But there's no path to world domination by selling carburetors (or bowling alley control systems) so those problems never get solved by anyone with the means to do so. It's a shame, really.

> retrofitting very old machine tools with modern motion controls.

I have a friend who has automated an old food production line. Eg peanut butter making, Muslie roasting etc.

With no background in programming he got machines hooked up measuring various things - eg peanut butter levels with ultrasound, roast temps, rotation rates.

Zero version control and eyebrow raising methods but massive improvement in safety and throughput. Really impressive.

I think you’re absolutely correct.

I suspect it is practical to do a small one-time run of maybe a 1000 copies of a purely hardware device and sell them. Of course, that is just kicking the can down the road a bit farther… At least dealing with obsolescence with purely analog parts is simpler…

As soon as software appears, things become less ideal. Now the HW and SW become highly specialized to a single vendor’s single family of devices. Plus, bugs and internet connectivity almost ensure a potentially unbounded number of updates or forever broken functionality. Even toolchains fade into obsolescence…

Sure, one can redesign for a new microcontroller from a new vendor. Maybe with new RTOS too. Of course, that is almost as much work as the initial design and probably done by someone else at a later date. Yuck!

(And I saw this as a software engineer!)

Definitely something to be said of the old habit of a simple analog design with schematics fully provided to the end user. This is something probably still needed today, but perhaps commercially challenging, especially with countries like China in play…

> There are many opportunities to retrofit old systems of all types with modern low-cost embedded technologies.

I'd leave out the "old" there.

An example: Friends work(ed) as engineers in theaters and music-venues. The costs for (DMX) equipment is insane. Some lamps, rigs or controllers are so expensive that shows only rent them. While there's lots of cases where you don't want to cut costs¹, quite often a cheap variant cuts it just fine. So apparently there's a lot of Pie's, Arduinos, ESP32s and old androids used to handle stuff. Custom, hacked, duckt-taped solutions. All of them are certain a startup offering generic but cheap modules that engineers can customize, would be an instant hit. Especially if "service" can be bought with it.

Or a friend who installs (climate) control systems in (food)factories, who cludges together arduino's and some hacky web-interfaces, who is certain that the industry would love some "off the shelve" solutions that installers can build their own hardware- and user-interfaces on is a very much wanted product. The industry (in EU) is according to him, one built on "lock ins", where a factory starts off with AcmeInc controllers and then slowly Acme will start charging almost "extortion-fee" alike figures just to fix stuff on a sunday night.

Or, a friend who has a fruit farm with lot's of freezer-storage (running hundreds of KWs), whith whom I cludged together some Arduino's and home-automation to control the freezers to run when electricity is cheap or free. Stuff like this is available for consumers at reasonable prices, but the moment the Amperages go to industry-level, the costs explode. Where a simple remote controllable power-switch can cost €1000s. I'm certain that there's hundreds of thousands of small warehouses, supermarkets, traders, logistics that could save a few hundreds to thousands of €'s a year, by controlling their freezers, heaters, pumps, pressurizers, dryers and whatnots smarter with "turn key" modules built from generic low cost hardware.

--

¹ relevant: There's a lot of "Nobody Ever Got Fired for Buying IBM" behind it: if you go for the cheap €800 chinese knock-off and mid-show something starts failing, its your fault. But if that €10.000 controller starts failing, it's no-ones fault.

Yup. I live on a river. We flooded catastrophically in 2019, which made me suddenly care about the government’s network of river level meters.

Turns out, they abandoned most of them in 2008, as it was the local council who were meant to maintain them, but they couldn’t afford it - a chat with the president revealed they’d been paying €15k a year for each station to maintain it and they’d cost about €100k each.

I went and looked at the battered remains. Sounding tube. Ultrasound transducer. Box of 90’s electronics.

Anyway, long story short I banged out six of them for them for about €80 a piece - the expensive bits were the batteries and the dinky solar panel. ESP32, ultrasonic transducer, GSM and LoRa (as two are outside of the cell network), haven’t had to touch them in five years and we now know when it’s time to evacuate, rather than wading out in the night while trees sail past us at 60kph.

So much stuff like this.

I do retrofits and I'll just assume the equipment he was using didn't have an analog encoder option or was ignorant of such equipment (likely using some cheap hobby level CNC thing.) I've worked with a few motion control systems such as ACS and Aerotech and they provide inputs for just about every encoder type imaginable. I retrofitted a CNC with 1500W 90V servo motors with analog resolvers using an Aerotech system, zero mechanical upgrades.

> Thank you for posting this, it reaffirms what I've been thinking for so long. There are many opportunities to retrofit old systems of all types with modern low-cost embedded technologies.

At my shooting range, where security is very important, we're replacing the old antiquated "fire stop" (when everybody on "shared" lanes has to stop shooting for there's at least one person going to change the targets) system with a modern system of electro-magnets and embedded devices. Mechanical safeguards where, by default, the 10 lanes (there two times 10 lanes, the 50 meters ones and the 100 meters ones) have to stop shooting and then an embedded system with a little controller for each lane that needs to be working to give the greenlight / "fire frei" ("free to shoot").

We mostly decided on the switch after a failure of the old system: the "free to shoot" system has recently been given even though there was still one person outside, changing the targets (nothing bad happened but this is a serious incident).

We asked for quotes for a system that is always "door locked" / red light / "fire stop", unless everything works fine and every lane gives the greenlight (each shooter has now one badge).

Same thing as in TFA: cost in hardware is minimal compared to the quoted prices. Like basically 1/50th of the quoted price.

[dead]