The end of PLCs?

Siemens software controller IS running under Linux.


Fair point. I don't know about the soft PLC itself, but I'd be shocked if the hypervisor wasn't something linuxy.



My point was that the box can still have windows, and windows issues don't affect the PLC.
 
I could save a ton of $ by replacing a lot of small PLC with remote IO but I wouldn't. You have to critically analyse your process and figure out the consequence if you you lose comm. If it's not a big deal, then sure, go ahead.

Also, along the same thought process. The "very pricy" PLC could seems like a drop in the bucket depend on the risk.
 
I had my first experience with DCSs recently as part of a DeltaV system. I only built a panel though. Nothing on the software side of things. Isn't there still a physical controller with that system or am I wrong?

Edit: I guess AB remote IO is the same concept but there is still a controller involved.

I did a cabinet design with DeltaV CHARMS setup a few years back and as I remember, it has redundant controllers. Redundant power supplies and ethernet . Looked pretty slick, like you though I was not involved with the software side. Would have liked to get my eyes on that enviroment.
 
I'd argue that there really isn't a fundamental difference between a PLC, PC, Tablet, Phone, etc anymore. Microprocessors and memory have become so cheap, the real difference are in the software and intended application. The big difference in industrial hardware is mainly ratings for the harsher working environment. There are slower development cycles and longer life cycles with industrial hardware, but it appears that trend is going away as well since, most things are done in software now and hardware component availability life cycles are so short.

There is no way a 1756-L72 will be supported as long as a PLC5 was, but because most everything is complemented in software, its easy to migrate to a 1756-L82 or future versions which has completely different hardware.
 
The nearest thing to a 'PC' based industrial controller I have seen lately, is the GRV-EPIC-PR1 Edge Programmable Industrial Controller from Opto 22.
You can use their own flow chart based programming environment, PAC Control, or if you prefer use the built-in CODESYS. Apart from other things it supports such as MQTT, Node-RED, you can also choose to use the Secure Shell Access, which is a Linux based OS.
Has its own rack mount IO, and/or will talk to remote IO using Modbus/TCP, OptoMMP, Ethernet/IP, to name a few!!!
 
What changes between your scenario and mine? A server in a closet somewhere vs a PLC in a closet somewhere? You won't be able to replace the performance of a $10k PLC with a $200 PC. Maybe with a $3k PC, but how often do you see a typical PC running for 20-30 years?

..and if you do, at what ULTIMATE cost LATER?

I was on a site a few years ago where a key piece of equipment in this small plant was running on a DOS computer, fitted with OPTO Brain Boards.
I was told that "..a key person in management (now) was a budding technician (then), and built it from scratch..." Unfortunately, their budding expert simply didn't document everything, and, moreover, they were one failed-20-year-old-hard-drive away from a closed plant.

I was on a different site, where they had a 'really smart' intern one summer, who solved a tricky problem using Arduino boards. Yes, it worked great. So, I asked the plant manager: "Did your intern leave behind the source code, and instructions on how to build a PC to actually run and download the code...as well as spare boards to match what he put it?"
=silence=
Yeah, we old-timers have heard all about the "PC-based PLC".

Haven't seen one work out yet.
 
Generally speaking, I do think that PC based PLCs are coming more and more.
But I do expect that product life cycles will be shorter, and reliability will be less.

I'd argue that there really isn't a fundamental difference between a PLC, PC, Tablet, Phone, etc anymore. Microprocessors and memory have become so cheap, the real difference are in the software and intended application.
I dont agree with that. In PLCs there is a difference in both hardware design and testing that emphasize ruggedness (can take a beating mechanically and electrically), reliability (wont last only 2-3 years like consumer electronics do these days), and simplicity (not being dependent on for example USB drivers for the devices used).

[..] it appears that trend is going away as well since, most things are done in software now [..]
True, but even for software the big PLC brands are still much more conservative and perform more testing than consumer/office brand products. In particular for the devices that control something.

[..] and hardware component availability life cycles are so short.
The availability of components in current PLC designs cannot expected to last 20+ years like they used to. But for current PLC designs the availability of components must still be much longer than consumer/office grade equipment.
The other way to handle the obsolescence of parts is to always have a migration path to a newer version of the same hardware. The big PLC brands will not leave their customers high and dry when the parts are unavailable.

@jdbrandt. I think that what you describe is that people that are used to work with control systems and PLCs have a totally different mindset than people that work with PCs. We are used to think about things like doing proper backups and documentation because of hard-learned experiences. When a non-PLC guy (manager) asks another non-PLC guy (intern) to make a control system, it is a disaster going to happen.

edit: I just reread my own post and realize that my words sounds denigrating towards PC programmers. Of course PC programmers also document things, but the mindset is different. The PC programmers i know of think for example that it is vital that all programs are stored on the same Git server, and the latest updates are always automatically pushed to the endusers.
 
Last edited:
Just to stir the pot a little, Valmet/Metso has been using industrial computers as the controller in their Valmet DNA/Metso DNA DCS for years. Their main markets are in pulp&paper and power plants, but also some amount of chemical plants.

To my knowledge the OS they are using is some version of Linux, although I haven't been involved with their system on that deep level that I could say for sure.
The system is used only for the process control, though. The safety side is done with 3rd party safety PLC.
 
Their main markets are in pulp&paper and power plants, but also some amount of chemical plants.
[..]
The system is used only for the process control, though.
"only" !
The costs of downtime or product damage is much higher than on a single machine.
In proces control the stakes are higher, even if you ignore safety, they usually run 24/7 with no downtime available for maintenance. So in proces control you usually have redundancy and configuration-in-run. Features that you normally have in higher-end PLCs or dedicated DCS systems.

So that does sound as a good reference for PC based controls.
 
But when Skynet creates it's mobile robots PLC's won't be needed.

A robot can set all the adjustments and turn everything on as needed and load and unload the product, logging each part and all runtimes, counts, etc.

And, don't have to worry about the PLC program re-writing itself after a while, either.

Come-on Skynet!
 
The danger of having PC instead of PLC is many people will be inside cabinets on facebook or plc.net, need to be careful of the upcoming dangers of social media :p
 
PLC/PC stability

PLC are stable and do not require updates, that makes them secure and isolated from malware. Can PC based do the same? Not for the price.

PLC's DO require updates and they are not ultrasecure (likely more than an windows PC, but not "secure"). They have vulnerabilities and can be attacked by malware. read your firmware release notes and you will see things that allude to this. Also consider there is a homeland security group that is dedicated to PLC and automation control security vulnerabilities due to the crippling economic effects a hack could create.
 
To my knowledge the OS they are using is some version of Linux


Maybe they have an in with Torvalds :ROFLMAO:


PLC are stable and do not require updates, that makes them secure and isolated from malware. Can PC based do the same? Not for the price.


PLC's are more security through obscurity. There are alot more PC's in the world which makes them a more common target.
Speculation: Alot of hackers don't know how to program a PLC, let alone that they even exist.



I've seen more and more articles lately in PC press bringing focus to the fact that PLC's exist and are a viable target for hacking so it will probably be only a matter of time before we start seeing more and more ransomware attacks targeting the control systems.
 

Similar Topics

I'm on the hunt for for a "Balance of Plant" PLC (or alternative) that is capable of exchanging (sending and receiving) a small number of bit...
Replies
6
Views
2,156
I am bidding a controls design, conveyor-type situation, and one of my client's top priorities is cost savings. They have European AB and Siemens...
Replies
7
Views
4,416
Hi everyone, Now I am considering to use one out of Allen Bradley PLC products for my simple project. In my project, first the PLC receives the...
Replies
5
Views
5,515
Our punch press has a SLC500 with HELM weight module (HM-604-WM) for tonnage monitoring. The operator enters the weight range on the HMI, which is...
Replies
9
Views
139
I am converting a SLC 500 to a Compact Logix. I plan on using a Compact Logix 5380 with conversion Kit. The problem is that the analog input cards...
Replies
1
Views
114
Back
Top Bottom