Opinions wanted on the Open Souce PLC/Industrial Automation

With the Automation Directs Open Source Arduino Compatible PLC and the Open plc Project, plus all the Raspberry Pi HMI's, what are your opinions on open source automation? In an industrial environment, I don't think C++ is the way to go, and if it's done graphically it has to be better than the DirectSoft programming. I could imagine Version control and hardware revs turning into a nightmare.

Thanks,

https://open***************/

For some reason, I can't fix the URL to the Open PLC Project.

Open source has brought a lot of benefits to the PC programming world. The PLC world has picked up some of those, buried deep inside the FW, but most of the software we interact with is still very proprietary. I really hope something happens to help free things up, but I really really doubt that it will any time soon.

I'm all for open source. But this is the dystopian PLC world of reality


I'm not holding my breath. I have wanted to start working on an open source HMI solution in .Net for a long time, but that is a large scale project and I just don't have the time, between work, and small children. So my dream just sits there in my head collecting dust.

I am usually an advocate for using open source as much as possible, but for industry i have my doubts.
The main reason we keep buying from the big brands are reliability and support backed by an organisation with muscle$$.
Another factor is the clients prefered vendor.

I doubt an open source project DCS or high tier PLC will be able to perform all tests in hardware and software neccessary for critical control, especially for redundancy and safety.
Open source is mainly developed by people on their spare time like a hobby.

Then there are companies like Red Hat (now part of IBM) that make their money from support and costumizations of Linux, perhaps there will be some company that steps up to provide the same services for an open source control system.

I'm more of a beliver in open communication and import/export functions between brands. But they will never willingly open the fences between their walled gardens - that would put the customer in charge.

It's a tough call.


Corporate customers will want a vendor they can call; or sue. Either motivation is typically a showstopper that will move them away from open source projects. Another advantage to commercial systems are the availability of near-turnkey solutions, e.g. S7-1200s, MicroLogix, etc. They know they will have to program the logic for the process, they don't want to have to load the operating system as well, or, more importantly, maintain the in-house skill to do so. In then end they "cannot afford free software."



But they also want broken, or poorly designed, tools fixed, which will typically happen more slowly, or not at all, with a proprietary vendor who decides what to fix on a profit-based metric e.g. see recent posts #14 and #15 here:

• https://www.plctalk.net/qanda/showthread.php?t=124619

Meanwhile fixes typically happen more quickly on active open source projects, often within hours or even minutes of being brought to the project's attention. Because, although some are hobbies as noted earlier, it is reputation, often tied to ego, that is the motivator.


Also, an open source project can also bring tremendous and flexible resources to bear, such as individuals who have superior knowledge of specific domains but would only work on a given project for a few hours a year once the project is mature. Whereas a commercial concern has to do all work with the staff at hand, while paying for salaries and benefits, managing, HR, NDAs, etc.; it is not economical to get a specialist in to work short-term on a specific task.



This forum is akin to an open source project: I suspect the value of free advice regularly given away here would be enough to support an engineering firm that could easily bill at a very profitable rate on the open market.


Regarding version control, I consider Git/Github/Gitlab/etc. to have solved that adequately: the automation of testing and code coverage via tools like Travis is a thing of beauty.


Another place where open source wins big is innovation. Have any of you worked with something like Docker? I just started and it is amazing, and there are many related tools as well.



Hardware costs are an interesting topic. Open source seems cheap, with a RaspberryPI a tenth or even a hundredth of even low-end commercial PLC units. However, once things like I/O and terminal strips are added, and more significantly the labor to organize and assemble it all, maybe the commercial units are not that expensive after all.


So there are many benefits to open source, but rarely are they enough to override the primary economic motivations first mentioned above; so I am saying roughly the same thing as ojz0r before me. I cannot speak to the specific projects referred to in the OP, but I suspect they will work in some situations. I have been impressed with AdvanceHMI, but then I am comfortable in a VB environment. I did my first project with it recently; I had to write only a handful of lines of code to implement the desired interface. That seems to be a working business model for open-source: give most of the product away; charge for a few add-one, but not much; presumably they make it up on volume. It's a bit of a niche, and now VB is going away.


This has been a bit of a ramble; thanks for reading.



Here's that link, but it's broken:Click on it, and you will see summat like the image below; remove the "x" from the URL in the address bar (highlighted in red in that image; Firefox shown; YMMV) and hit Enter.


Or take the first hit here: https://www.google.com/search?q=open+plc+project

drbitboy said:

They know they will have to program the logic for the process, they don't want to have to load the operating system as well, or, more importantly, maintain the in-house skill to do so. In then end they "cannot afford free software."

Click to expand...

To be fair, a Hypothetical open source PLC project could provide ready to go boot images for common common hardware. All you would have to go is flash a SD card or such and away you go.

Though you would definitely want to use SLC SD card, for reliability but they aren't cheap.

drbitboy said:

Hardware costs are an interesting topic. Open source seems cheap, with a RaspberryPI a tenth or even a hundredth of even low-end commercial PLC units. However, once things like I/O and terminal strips are added, and more significantly the labor to organize and assemble it all, maybe the commercial units are not that expensive after all.

Click to expand...

Well yes, but that could be easily rectified by using existing IO Modules like for Automation Direct Productivity 1000, That now overs a Arudio Controller. (It needs a RPi Controller with provided codesys drivers)


I also know of another company that is working on a low cost RPi based PLC that will run Codesys, but they haven't publicly announced yet to my knowledge.

drbitboy said:

So there are many benefits to open source, but rarely are they enough to override the primary economic motivations first mentioned above; so I am saying roughly the same thing as ojz0r before me. I cannot speak to the specific projects referred to in the OP, but I suspect they will work in some situations. I have been impressed with AdvanceHMI, but then I am comfortable in a VB environment. I did my first project with it recently; I had to write only a handful of lines of code to implement the desired interface. That seems to be a working business model for open-source: give most of the product away; charge for a few add-one, but not much; presumably they make it up on volume. It's a bit of a niche, and now VB is going away.

Click to expand...

Just switch to C# it's very much alive and growing. I love C#. You can use AdvancedHMI in a C# project even if the library code is VB.Net.

AdvancedHMI has so much potential but the problem is that Archie the Dev doesn't want to invest the time to modernize it. It needs to be ported to a cross platform UI framework. it currently uses WinForms which while maintained is a very old code base and is Windows only. The other reason he doesn't want to do it is because none of the more modern UI frameworks provide a complete Visual Designer, and he wants people that don't know anything about programming to be able to use it. I think that misses the core users base though that is more advanced users.

theColonel26 said:

...You can use AdvancedHMI in a C# project even if the library code is VB.Net.

Click to expand...

Good to know, I suspected that was the case.

theColonel26 said:

... AdvancedHMI has so much potential but the problem is that Archie the Dev doesn't want to invest the time to modernize it. It needs to be ported to a cross platform UI framework. it currently uses WinForms which while maintained is a very old code base and is Windows only. The other reason he doesn't want to do it is because none of the more modern UI frameworks provide a complete Visual Designer, and he wants people that don't know anything about programming to be able to use it. I think that misses the core users base though that is more advanced users.

Click to expand...

Interesting point: if it is made more powerful, and therefore more complex (the slope is always positive between those two), to gain members of the cognoscenti, that will lose some of the great unwashed; there are almost always more of the latter. It depends who is really that "core" user group that best serves the business model. It is certainly true now that users do not need to know about programming: useful displays, including scaling, interaction, etc., are straightforward to implement via simple visual controls (text boxes and such) without touching a line of code in the .vb files.

Openplcproject

openplcproject dot com

weird that you can not post the whole URL without it being censored.
open***************

drbitboy said:

.


Corporate customers will want a vendor they can call; or sue.

Click to expand...

this is an important point. If I mess up programming and it kills someone, then liability will fall on me. If a PLC manufacturer has a flawed design that causes outputs to fail and it kills someone, then they will get brought into the mix and my insurance breathes a sigh of relief. If I design a system with a rasberry pi, and one of the channels fails because of a design flaw in the pi, rasberry is gonna say, no way buddy, this was not intended to be used in a industrial application.

diat150 said:

this is an important point. If I mess up programming and it kills someone, then liability will fall on me. If a PLC manufacturer has a flawed design that causes outputs to fail and it kills someone, then they will get brought into the mix and my insurance breathes a sigh of relief. If I design a system with a rasberry pi, and one of the channels fails because of a design flaw in the pi, rasberry is gonna say, no way buddy, this was not intended to be used in a industrial application.

Click to expand...

This right here. I was trying to find the words to say that in my post above, and ended up deleting it because my words sounded dumb.



You said it better and more concisely than I could.

I would also like to point out what a lot of people here have really seemed to missed.

There is a HUGE range of PLC products.


There is anything from 10 Million dollar continues process lines that have the capacity to not only destroy themselves if timings are messed up, but to kill and maim people, to a $3000 little table top machine that just has lights and sensor to tell the human assembler what to do next.

I think there is a large market for Open Source PLC stuff in the Low Risk, High Volume, Low Cost area.

Lot of things that are currently embedded with custom hardware, but barley have the volume to justify the custom hardware would be much better off with a cheap opensource PLC, it would also allow there designs to be a lot more flexible.

Originally posted by theColonel:

I think there is a large market for Open Source PLC stuff in the Low Risk, High Volume, Low Cost area.

Click to expand...

I definitely agree with this. Users are always asking for low-cost plcs to do this or that but you often have to give up a good chuck of capability to get it. Open Source on cheap hardware would fit in this gap very nicely.

And, as I've always said, most of us seem to undervalue our time while overvaluing the time of others. Putting in a little extra effort to figure out how to use something less expensive seems to be right up our collective alley.

Keith