Merging modern software development with electrons and metal
Random header image... Refresh for more!

Automation Trends I Want To See

Here are some factory automation trends I’d like to see:

  1. Much more adaptation and appropriate use of modern software development techniques such as version control, design patterns, functional programming, unit testing, automated testing, model based design, hierarchical state machines (state charts),  and agile software development.
    1. From what I can see, most automation developers still aren’t borrowing from the best of mainstream development techniques.
    2. A big reason I started blogging was to promote these techniques.
  2. Use of modern programming languages.
    1. There’s been some progress here with more use of PC’s in automation and companies such as Beckhoff supporting the use of Visual Studio.
    2. There’s been some progress on the PLC front: IEC61131 isn’t remotely comparable to .NET, but it’s a big improvement over the past.
  3. Fewer poorly designed, vendor-specific programming languages.
    1. This is a pet peeve of mine, since I’ve had to deal with way too many.  Modern ICs have advanced far enough that we shouldn’t have to keep reverting back to 1970’s-style programming languages.  If nothing else, please use IEC61131 with PLCOpen for motion control.
    2. Unfortunately, this problem isn’t going away, but at least with companies such as CoDeSys widely licensing IEC61131 software, it hasn’t gotten worse.
  4. Advanced automation programming books.
    1. The basic problem is there aren’t any.
    2. There’s a chance Frank Lamb of Industrial Automation: Hands On fame will improve the situation.
  5. Widespread kinematics in motion controllers.
    1. This really is a trend, especially in controllers targeted towards the packaging industry, and it’s a good thing: we have enough cheap computing power that we don’t need to treat linked axes as a bunch of unrelated axes anymore.
  6. “Everything” on industrial networks (fieldbus and Ethernet) at a reasonable price premium.
    1. The situation is getting better, except for the price premium.  I’ve recently seen more unusual networked components such as barcode readers, smart cameras, and laser distance sensors.
    2. Of course, if you’re trying to get all these devices to work on the same network, your options shrink dramatically.
  7. “Everything” on industrial real time networks (fieldbus and Ethernet – such as CANOpen, EtherCAT, and Ethernet PowerLink) at a reasonable price premium.
    1. Other than servo drives, there are a lot fewer components available for the real time networks.
  8. Fewer standards, but still some variety.
    1. It’s hard to have one standard that adequately meets all the varied needs of the wide world of automation.  For example, in Ethernet there are major tradeoffs between use of standard components, robustness, and real time capabilities.
    2. But  there are way too many standards.  In Ethernet, 2-3 industrial standards should be plenty.  I think we’ll have to live with 5-6 (Ethernet/IP, Modbus TCP, Profinet, EtherCAT, Ethernet PowerLink, and maybe CC Link IE, which is sort of Ethernet (PHY only), which is manageable, but there are at least 20.  And too many companies still invent their own serial protocols (RS232, RS485, CAN, etc) instead of using a standard.
    3. I guess all these standards is good for the makers of gateway devices…

End of rant!

In between my Robot Primer posts, I will highlight some cool networked devices.


There are no comments yet...

Kick things off by filling out the form below.

Leave a Comment