Ignition Software - questions

I'm new to this process. I've been tasked with researching, planning and ultimately implementing an MES solution.
We are a manufacturer in the automotive industry. We have about 100 stations.

- How long does it take to fully implement Ignition? The FAQ says it's "ridiculously easy to install". I can't find information about roll out time and resources required to successfully launch.

- What are some limitations?

- Has anyone attempted and ultimately abandoned Ignition, if so, why?

- Are their other questions I should be asking?

- Does anyone know of any comparison type charts/reviews for MES solutions?

Any help is greatly appreciated!

Welcome to the Forum !

Much depends on how you define "fully implemented", right ? While getting "up and running" with the servers and the first few tags is easy, an automotive MES system is a multi-month project.

I do agree that Ignition is much faster and easier to install and deploy than any other HMI software I have used. Starting from a Linux virtual machine to getting my first HMI tags updating was less than an hour the first time I tried it. I've heard people cite similar under-an-hour installation times for their first attempts at Ignition on Windows computers.

Compared to the two solid weeks I'd spent updating and patching FactoryTalk View 4.0 on some Windows 2003 Servers previously, it almost made me cry.

The limitations I ran into with Java are also the reason I abandoned Ignition in my company.

This is very specific to our company and its history; we rely on Python and its NumPy and SciPy libraries extensively. I like to say we have a "mathematician in captivity", whom we keep in a nice dark quiet air-conditioned room padded with Python reference books.

But Ignition's Python scripting and extensibility is based on the Java version of Python, which doesn't have NumPy or SciPy. And when people say "just get the source code and write your own in a different language", I have to tune them out.

Inductive Automation had some recent trouble with their Rockwell Automation drivers, so if that's your platform for PLC control, then you're fortunately over that hump.

Have a sincere talk with your IT department about how they deploy, update, and secure Java. That's the conflict that I've heard other people encounter; they could run Ignition fine if the IT department hadn't taken away Java support on their clients.

tremes said:

- How long does it take to fully implement Ignition?

Click to expand...

Depends on what you consider "fully implement" to really mean.

I could be up and running in less than 20 minutes. That time includes the actual download of Java and the Ignition installer.

Depending on what you actually need to configure for your specific environment, that will add time. 1 PLC? 100 PLCs? 1 Database? 100 Databases?

tremes said:

- What are some limitations?

Click to expand...

Ken touched on this already, but I think Java roadblocks will be present before Ignition roadblocks. But all depends on what you really want to do. I think you have to find very specific instances where a limitation will be difficult to get around, much like Ken's situation.

Ignition doesn't have a true Historian, it simply logs to a database w/o any compression(arguments for and against compression are easy to google). While you can do a lot with charts/trending in Ignition, they aren't as advanced as what you might find on a true historian trend. So depends on what you actually need, and how you actually want to present data.

tremes said:

- Does anyone know of any comparison type charts/reviews for MES solutions?

Click to expand...

Can't say I know of any, most will simply be marketing BS from all the manufacturers.

Simply getting the 'unlimited' licensing package is great because you can simply "get to work". I can't imagine how much time I've wasted designing/engineering around licensing limitations, all to manage upfront costs while the additional labor costs to do so seem to come out of my "free" time.

Having reporting in the package is great too, much easier than dealing with SQL Reporting, or some other reporting package.

At the end of the day, you probably need to write a User Requirements Specification to help you shop the packages to ensure they will meet your needs.

Paully's5.0 said:

Depends on what you consider "fully implement" to really mean.

I could be up and running in less than 20 minutes. That time includes the actual download of Java and the Ignition installer.

Depending on what you actually need to configure for your specific environment, that will add time. 1 PLC? 100 PLCs? 1 Database? 100 Databases?


Ken touched on this already, but I think Java roadblocks will be present before Ignition roadblocks. But all depends on what you really want to do. I think you have to find very specific instances where a limitation will be difficult to get around, much like Ken's situation.

Ignition doesn't have a true Historian, it simply logs to a database w/o any compression(arguments for and against compression are easy to google). While you can do a lot with charts/trending in Ignition, they aren't as advanced as what you might find on a true historian trend. So depends on what you actually need, and how you actually want to present data.



Can't say I know of any, most will simply be marketing BS from all the manufacturers.

Simply getting the 'unlimited' licensing package is great because you can simply "get to work". I can't imagine how much time I've wasted designing/engineering around licensing limitations, all to manage upfront costs while the additional labor costs to do so seem to come out of my "free" time.

Having reporting in the package is great too, much easier than dealing with SQL Reporting, or some other reporting package.

At the end of the day, you probably need to write a User Requirements Specification to help you shop the packages to ensure they will meet your needs.

Click to expand...

One quick comment, Ignition does have a more traditional historian type module (Called the Tag Historian module). It does compression while still using SQL as the backend.

I am a very big flag waver for Ignition and love the product.
I did attempt an OEE live demo for a dog and pony show for a potential customer.

The MES portion of Ignition can be difficult to set up.
There is a dedicated training course in Folsom CA just for MES.
I have not checked Inductive Univerisity, but there are videos for MES

Ken, Paully's, Dravik and CurlyandShemp,

Thanks to all of you for taking the time to respond. This is all great information. I'm going to review the information and respond with some more specific questions.


:site:

Ken Roach said:

Welcome to the Forum !

Much depends on how you define "fully implemented", right ? While getting "up and running" with the servers and the first few tags is easy, an automotive MES system is a multi-month project.

Click to expand...

This is great information for me to start with. My boss is sitting in his office expecting that I'm going to sign up for an account at Inductionautomation.com, download the software (because their website says it's "ridiculously simple to install and use"), download the MES app and we're good to go.

Essentially, after reviewing what this project is going to take, I'm working to build a case to help educate him and align his expectations with reality.

"fully implement" = installing, configuring automotive MES, using ignition to help drive quality and process improvements.

Ken Roach said:

The limitations I ran into with Java are also the reason I abandoned Ignition in my company.

Click to expand...

Can you speak more to this? Java was a flag for me because of its dwindling status as a programing language. I know they aren't going to pull the plug on it. However, I can't help but question the fact that Ignition is built in JAVA. There will always be this, "what happens if"?

Ken Roach said:

Inductive Automation had some recent trouble with their Rockwell Automation drivers, so if that's your platform for PLC control, then you're fortunately over that hump.

Click to expand...

Are there any other such troubles/issues I should investigate?

Ken Roach said:

Have a sincere talk with your IT department about how they deploy, update, and secure Java. That's the conflict that I've heard other people encounter; they could run Ignition fine if the IT department hadn't taken away Java support on their clients.

Click to expand...

I plan on this. I want and need these guys on board.

Personnel resources are stretched thin as it is. Should I be investigating the need for added head count for deployment, set-up and management of Ignition?


Thanks Ken - your response was greatly informative.

Paully's5.0 said:

Depends on what you consider "fully implement" to really mean.

I could be up and running in less than 20 minutes. That time includes the actual download of Java and the Ignition installer.

Depending on what you actually need to configure for your specific environment, that will add time. 1 PLC? 100 PLCs? 1 Database? 100 Databases?

Click to expand...

This is going to take a while for me to investigate. Is there a rough, rule of thumb I could go by? Something like 10 databases = XX time and 100 PLCs=XX time?

Paully's5.0 said:

Ken touched on this already, but I think Java roadblocks will be present before Ignition roadblocks. But all depends on what you really want to do. I think you have to find very specific instances where a limitation will be difficult to get around, much like Ken's situation.

Click to expand...

What I want to to is call someone in for a turnkey solution!
Being that we are only in the investigation stage, *my* current challenge is to accurately establish scope and expectations for all parties involved.

Paully's5.0 said:

Ignition doesn't have a true Historian, it simply logs to a database w/o any compression(arguments for and against compression are easy to google). While you can do a lot with charts/trending in Ignition, they aren't as advanced as what you might find on a true historian trend. So depends on what you actually need, and how you actually want to present data.

Click to expand...

I'm adding this to my list for review later. Thanks!

Paully's5.0 said:

Can't say I know of any, most will simply be marketing BS from all the manufacturers.

Click to expand...

Ok, it's not just me. There is NO shortage of marketing BS out there (that's what I'm finding)

Paully's5.0 said:

Simply getting the 'unlimited' licensing package is great because you can simply "get to work". I can't imagine how much time I've wasted designing/engineering around licensing limitations, all to manage upfront costs while the additional labor costs to do so seem to come out of my "free" time.

Click to expand...

I'm going to add this to my project scope. I'm very worried that we are going to be under-resourced for this project. The lack of resources, if fear, are ultimately going to lead us to failure.

Paully's5.0 said:

Having reporting in the package is great too, much easier than dealing with SQL Reporting, or some other reporting package.

Click to expand...

Paully's5.0 said:

At the end of the day, you probably need to write a User Requirements Specification to help you shop the packages to ensure they will meet your needs.

Click to expand...

This is my current mind set. Based on the limited research I've done, the people around here have VAST differences in expectations. Some people think it will be impossible to implement, others think it's as easy as programing the microwave to reheat your coffee.

Thanks for your detailed information!

Dravik said:

One quick comment, Ignition does have a more traditional historian type module (Called the Tag Historian module). It does compression while still using SQL as the backend.

Click to expand...

Thanks - I'm going to investigate historian information in the near future.

curlyandshemp said:

I am a very big flag waver for Ignition and love the product.
I did attempt an OEE live demo for a dog and pony show for a potential customer.

Click to expand...

Why do you say "attempt"?

curlyandshemp said:

The MES portion of Ignition can be difficult to set up.

Click to expand...

Could you provide examples of the difficulties with set-up. These would be gold to me right now. It would really help me while defending my project scope and expectations with management.

curlyandshemp said:

There is a dedicated training course in Folsom CA just for MES.

Click to expand...

Can one deploy a functional and usable MES project without going to the course? Are other courses needed?

curlyandshemp said:

I have not checked Inductive Univerisity, but there are videos for MES

Click to expand...

I'll check into this.

Thanks CurlyandShemp!

Inductive University is outstanding. I have used it as my sole training method to develop SCADA applications. The videos are excellent and they include online testing and keep track of your progress.

I would encourage you to sign up for a trial, register with IU, and watch some of the video tutorials. I have not been pestered by salesman by registering. They will email you and may call you once every couple of months to check on you if you end up purchasing a product. My experience with Inductive Automation has been stellar, although I am still a rookie.

I have no experience with MES, but it is a whole separate "school" in their library of course:

https://inductiveuniversity.com

tremes said:

This is great information for me to start with. My boss is sitting in his office expecting that I'm going to sign up for an account at Inductionautomation.com, download the software (because their website says it's "ridiculously simple to install and use"), download the MES app and we're good to go.

Click to expand...

Oh...boy, you're in for a ride aren't you?

Here's the deal, no matter what solution you decide to use if you don't know where you want to "go" with it, your are prime for failure. Selecting an MES package is only 1% of the project.

Anyway, Ignition is very easy to use and if you have experience with any industrial software you'll pick up Ignition and may not even need the training. Personally I never attended the training but easily become certified with the standard Ignition product (not MES ). If you are new to MES in general, and new to MES software, you should absolutely go to the training as it will pay for itself. Personally, if I had an MES project I would go to the training.

Otherwise just find an integrator to do the work for you. Which in this case is probably what you need to be doing. Of course before you track down and integrator you have to create some type of user requirements as I've said before. Otherwise how do you even know what you're trying to buy? What is the budget? I hope you've got $100,000+ to spend.

Paully's5.0 said:

Here's the deal, no matter what solution you decide to use if you don't know where you want to "go" with it, your are prime for failure. Selecting an MES package is only 1% of the project.

Click to expand...

+1 to what Paully said.

There's a Proverb that says "Where there is no vision, the people wander aimlessly". Other versions say "the people perish". Biblical references aside, that's also true with projects. Try to get them to share a vision, and the rest can flow from that.

That's not to say that there won't be a lot of changes as you go. Sometime the customer (in this case, the boss) needs to see something in-hand before he can say "It's not quite what I had in mind".

Oh! The Java limitation Ken is speak of that Jython (the Java implementation of Python) cannot use C-Python libraries. Numpy and Scipy fall into this category.

While Java itself seems to be falling off a bit, the JVM has not, and it's used in other languages. Scala comes immediately to mind.

tremes said:

Why do you say "attempt"?


Could you provide examples of the difficulties with set-up. These would be gold to me right now. It would really help me while defending my project scope and expectations with management.



Can one deploy a functional and usable MES project without going to the course? Are other courses needed?




I'll check into this.

Thanks CurlyandShemp!

Click to expand...

Attempt, meaning my live Dog & Pony show did not go well with customer. Probably due to my lack of understanding how the 'canned' OEE project worked from Inductive Automation.

It may be my lack of working knowledge about OEE that made it difficult for me to set up. Perhaps a more experienced OEE programmer may not have had the difficulties I had.

Up until Inductive University was released, the documentation s u c k e d . Inductive were aware of that and even hired a technical writer a few years back to try and improve the documentation.

Before IU, the documentation merely stated but did not actually explain features. Post IU, any time I have a memory lapse or need to learn a new feature I go to IU and watch the respective video(s).

OkiePC said:

Inductive University is outstanding. I have used it as my sole training method to develop SCADA applications. The videos are excellent and they include online testing and keep track of your progress.

I would encourage you to sign up for a trial, register with IU, and watch some of the video tutorials. I have not been pestered by salesman by registering. They will email you and may call you once every couple of months to check on you if you end up purchasing a product. My experience with Inductive Automation has been stellar, although I am still a rookie.

I have no experience with MES, but it is a whole separate "school" in their library of course:

https://inductiveuniversity.com

Click to expand...

Thanks for the info. You call yourself "a rookie". How long have you been at it? What sort of skill set did you start with?

I'm encouraged to hear your feedback about the IU!