Factory SQL/PMI comparisons

Can anyone provide info regarding how different this product is when compared to something like Citect or Wonderware.
Also, are there any Integrators in Victoria, Australia?

Sam – Probably best for me not to tout the benefits of IA software over WW or Citect other than the licensing model which I’m sure you know. I would say from an IT viewpoint of deployment across a wide network, it is pretty easy to deploy IA software. The functionality is all there with IA software – datalogging, history, alarming, batching and recipe management, downtime tracking, trending and reporting. How the system is designed is different. There is an integrator, Dennis Hayes based out of Queensland who also does work in Victoria. He is very familiar with Citect and Wonderware who you should speak with. His company is DMH Automation. You can look up his contact info on the IA website.

The design is more “database centric” and distributed (web/Java Launched) than other HMI systems.

It’s relatively de-coupled from the PLC by design, which allows a great deal of flexibility. Consider the difference between direct register addressing on a PanelView to a tag database of a standalone HMI - I would give the same analogy of transitioning from a standalone HMI to our distributed system. Each step adds a little complexity for other gains. Where FSQL/FPMI then differ from other distributed HMI systems is their fundamental use of SQL databases and Java Web Start technology. The first works with any SQL database, the second allowing clients to run without an “installation” or “configuration” process.

Sam you can call me on mobile from my web site and I would be happy to talk further regarding IA and SCADA’s.

thanks

Dennis

Thanks everyone,

Dennis I will give you a call once my work load drops off to discuss things
Regards,
Sam

I’m an end user replacing my existing Wonderware system with the IA product. Our Wonderware system has been in service for more than a decade. It was the best choice at the time. The technology was far more powerful, user friendly and cost effective than typical existing solutions. Wonderware, like most traditional SCADA systems, will have issues every time a new Microsoft Operating System is released. Although my current system works well, my version of Wonderware can’t be made to run under Vista and would require a costly upgrade. I have therefore been looking at SCADA replacement choices for three years now.

The three things I like the best about the IA product is the use of web based technology, the openness of the database, and the technical support.

By using the web based technology, I can develop screens from any computer on the network without any additional software and more than one tech can develop at the same. The client software is just a web browser (nothing to purchase or configure) and any PC, even my notebook, can view and develop screens. This is a huge benefit when bringing a new system online or troubleshooting an existing system.

I like the openness of the database-- the idea that I can use any popular database, including open source. When a vendor embeds a database into their product, they may modify it, place tag limitations on it, and then charge more than its retail value. I believe that this is the case with InSQL. This also locks the user into one database supplier and usually whatever database version the SCADA provider is selling at that time. To change to a new database requires a SCADA software upgrade.

As to technical support, I have never seen any company that supports their customers as well. Several times I have been shown my configuration mistakes in a go-to-meeting session. After the mistake was found, the techs really went the extra mile and helped with other questions and provided some quick training. I have posted questions to the forum (even on holidays) and had an IA person answer in less than 10 minutes. I’ve even posted SQL database questions and received answers. IA didn’t sell me the database Microsoft did, yet they were the ones who helped me out.

Thanks for the info Henry.
What database are you using?
Any reason you did not use MySQL?

We are using an existing server MS SQL 2005. I think the MySQL is a reasonable choice. If I were going to use a new server I would most likely be MySQL. I think most of the IA technotes lean towards MySQL. If you want to talk to an end user you can call me. I’m in California, best of luck on your SCADA project.

Yeah either MySQL or MSSQL will work good. MySQL is good because it is widely deployed in web architechures. However, It is hard to scale when used in conjunction with IA, unless you are extremely proficient with Linux, and MySQL scaling techniques. I prefer MySQL regardless, and the scaling efforts are trivial once you know how to implement them.

Kyle,
excuse my ignorance but what do you mean by ‘scaling’?

Scaling is a term used to describe how a computer, program, or network (or all 3) handle increasing load (more users, more transactions, more data, etc), and strategies to improve performance under increased load.

See: en.wikipedia.org/wiki/Scalability

Hey Sam,

Scaling helps improve performance of an application. While vertical scaling(faster, more expensive hardware) is easy, sometimes it is cost prohibative. The more traditional way of scaling MySQL, replication, wont work with IA without a Proxy in the mddle. However, in saying that, IA is extremely good job at scaling their stack, so that scaling to more users isnt linear. Lets say you have 1 user, and it needs 25 queries per second to sustain the client. If you add another client, it wont need another 25 queries per second to sustain those 2 clients, it might need 30 queries total.

At one of my plants, we run 10 concurrent users, trending 500 datapoints every 5 seconds, and 100 every second. Our IA stack runs perfectly, and has never crashed or needed an unplanned shutdown.

If you have any other questions, blast me a private message and we can talk about scaling more in depth.

I just want to add, that 99.9% of all IA installs will never need to use scaling techniques as a single database instance ill normally get the job done with ease.