Some idiot with college degree in office no-where near the place sees that we have these PCs here. And then they go over compliance list and mandate this is needed. Now go install it and the network there...
Or they want to protect their Windows-operated lifts from very real and life threatening events like an attacker jumping from host to host until they are able to lock the lifts and put people lives at risk or cause major inconveniences.
Not all security is done by stupid people. Crowdstrike messed up in many ways. It doesn't make the company that trusted them stupid for what they were trying to achieve.
For the same reason people want to automate their homes, or the industries run with lots of robots, etc: because it increases productivity. The repair shop could be monitoring for usage, for adequate performance of hydraulics, long-term performance statistics, some 3rd-party gets notified to fix it before it's totally unusable, etc.
I have a friend that is a car mechanic. The amount of automation he works with is fascinating.
Sure, lifts and whatnot should be in a separate network, etc, but even banks and federal agencies screw up network security routinely. Expecting top-tier security posture from repair shops is unrealistic. So yes, they will install a security agent on their Windows machines because it looks like a good idea (it really is) without having the faintest clue about all the implications. C'est la vie.
But what are you automating? It's a car lift, you need to be standing next to it to safely operate it. You can't remotely move it, it's too dangerous. Most of the things which can go wrong with a car lift require a physical inspection and for things like hydraulic pressure you can just put a dial indicator which can be inspected by the user. Heck, you can even put electronic safety interlocks without needing an internet connection.
There are lots of difficult problems when it comes to car repair, but cloud lift monitoring is not something I've ever heard anyone ask for.
The things you're describing are all salesman sales-pitch tactics, they're random shit which sound good if you're trying to sell a product, but they're all stuff nobody actually uses once they have the product.
It's like a six in one shoe horn. It has a screw driver, flash light, ruler, bottle opener, and letter opener. If you're just looking at two numbers and you see regular shoe horn £5, six in one shoe horn £10 then you might blindly think you're getting more for your money. But at the end of the day, I find it highly unlikely you'll ever use it for anything other than to put tight shoes on.
I imagine something keeps monitors how many times the lift has gone up and down for maintenance reasons. Maybe a nice model monitors fluid pressure in the hydraulics to watch for leaks. Perhaps a model watches strain, or balance, to prevent a catastrophic failure. Maybe those are just sensors but if they can’t report their values they shutdown for safety’s sake. There are all kinds of reasonable scenarios that don’t rely on bad people trying to screw or cheat someone.
None of these features require internet or a windows machine, most of them do not require a computer or even a microcontroller. Strain gauges can be useful for checking for an imbalanced load, but they cannot inspect the metal for you.
In my office, when we swipe our entry cards at the security gates, a screen at the gate tells us which lift to take based on the floor we work on, and sets the lift to go to that floor. It's all connected.
Remote monitoring and maintenance. Predictive maintenance, monitor certain parameters of operation and get maintenance done before lift stops operating.
It's a car lift. Not only would it be irresponsible to rely on a computer to tell you when you should maintain it, as some inspections can only be done visually, it seems totally pointless as most inspections need to be done manually.
Get a reminder on your calendar to do a thorough inspection once a day/week (whatever is appropriate) and train your employees what to look for every time it's used. At the end of the day, a car lift on locks is not going to fail unless there's a weakness in the metal structure, no computer is going to tell you about this unless there's a really expensive sensor network and I highly doubt any of the car lifts in question have such a sensor network.
Moreover, even if they did have such a sensor network, why are these machines able to call out to the internet?
These requirements can be met by making the lift's systems and data observable, which is a uni-directional flow of information from the lift to the outside world. Making the lift's operation modifiable from the outside world is not required to have it be observable.
