Myth#6 – No Need for DeviceNet Gateways - DeviceNet is Dead (and Profibus too)
I made myself famous (infamous?) a number of years ago by pronouncing DeviceNet dead. It wasn’t one of my best moments. I was enamored with Ethernet and thought that Ethernet would quickly take over the world of Industrial Automation. Well, I was wrong. Seems I am wrong almost every day about something. I was a real bonehead on this one.
A good story came out of it though. I got a call from a Marketing Manager over at this 2 billion dollar industrial automation company. It was frothing at the mouth mad at me. I didn’t care much. They weren’t a customer and had bought anything from me in years. Why he thought I would care is beyond me. Here’s the punch line: “John, do you know how many customers of ours are calling asking us what our Ethernet strategy is? We’re spending all day talking to these customers.”
I guess that’s why I’m not running a two billion dollar operation. I always thought it was a good thing to talk to customers. I want to know what their thinking is. What their manufacturing strategy is. What kind of problems they have. [Note to self: Talk to Jeff about disconnecting the phone in my office]
In any case, I realize now that nothing changes all that much and all that fast in Industrial Automation. It’s not a fast moving industry. We make things and some of the basic things, like Flour, Medicines and a whole host of other things, won’t change much over ten or twenty years time.
If you’re using a network like an AS-i Bus, a DeviceNet, a Profibus DP or something low level like that those things are going to be around forever. They’re perfected. We all know how to implement them, troubleshoot them and maintain them. They’re low speed and perfect for a lot of the jobs we need doing.
And since we have those networks there are a whole host things we need to move in and out of them. Stuff like barcodes, tag data and weights from scales. And that’s where gateways like the RTA 435 and 460 lines come into play.
Myth#7 – Ethernet is Best
No, it’s not the best. Properly sized, correctly implemented, smartly architected and well maintained it does a good job. But inherently it is not better than DeviceNet, Profibus or anything else. A lot of Engineers look at this completely wrongly. You should never start with the technology. Start with the problem. We have these kinds of devices. They use this kind of power. They are located geographically like this throughout our facility or around this machine. We need this kind of repeatability, this kind of determinism, this kind of response. The majority of devices we have support these networking technologies. Now, in light of all that what makes the most sense for this automation application? That’s how you approach the problem. Not with “Gee we all like Ethernet here so we’re going to do an Ethernet application.”
I hear a lot that certain companies are all DeviceNet. Or All Profinet or all Nestles Crunch or other such nonsense. I think it’s ridiculous. There are excellent reasons for AS-i as there are excellent reasons for all the others. I think it is very shortsighted to short circuit the analysis and jump to a technology.
Myth#8 – Serial is Dead
This is along the lines of DeviceNet is dead. Cost is the key factor here. Serial will always be the least expensive and lowest common denominator for lots of devices. If you’re a mechanical engineer [please stop reading you gear heads] and you are the project manager for the new Glue machine you’re likely to have skipped the need for comms and just figured we’ll give the customer a serial port. And for the hell of it, you going to invent your own protocol. Something called AMY where commands look like $AMYx. The AMY, of course, being the old high school girlfriends name and x being the command number.
I jest a little bit here but that’s what happens more often than not. Lots and lots of device manufacturers are driven by mechanicals and they outsource their electronics so it doesn’t have a place in the internal design process. You end up with crap like $AMYx and have to use a gateway to convert it to something useful.
BTW (means by the way for you non-texters) our 460 line is excellent for this kind of work. We are able to easily program it to poll these kinds of devices and turn the data into something useful for a ControlLogix, PLC5E, a GE90/30 or pretty much anything else.
Myth#9 – Wireless Won’t Work
I have heard this a lot. In fact, I’ve said it a lot. I think that I’ve been proved wrong here. There are lots of reasons and lots of places where wireless makes a lot of sense. Depending on how long the machine was going to stay in production I’d probably stay with wired systems for long term projects but on a short term machine (think snack machine) – wireless would be my first choice.
Myth#10 – Gateways are all alike!
No – not even close. There’s pretty much two classes of gateways. RTA’s and everyone else’s. All the other pack hundreds of features into a small box and let you figure out how to get the one thing you want it to do to work. RTA packs nearly nothing into our gateways so there’s almost nothing that you have to do to figure it out. It’s a big difference.
Myth#10.5 – Customization is Expensive
It usually is. Except that the smart guys that I’ve hired (that’s pretty much everyone I’ve hired, especially the females) have figured out a way where we can get custom projects done really quickly and pretty inexpensively. We do a lot of converting ASCII scale data to Floating Point, implementing custom message / response protocols and formatting, converting and totalizing data. We relieve the PLC of doing the processing which is so cumbersome in a PLC.
So, that’s my 10.5 myths of Device Converters. As always, please use the Contact Us page to send me any comments or thoughts about the things you see here. And remember the RTA staff has no control over the way I run my mouth off. The only thing they are allowed to do is to roll their eyes when I’m not looking.
Keep those card and letters coming!