PTO PROFIblog

So an engineer walks into a theme park…

2008-06-30T20:02:00Z

So an engineer walks into a theme park and while the kids are waiting in line for the kids’ coaster he carefully analyzes the mechanical construction details and postulates how the controls work. Mechanically, it’s round tubular steel tracks, supported by round steel tubular columns. The columns are embedded in cement footers with a welded plate on top. An I-beam is welded to the plate and the track is bolted to the beam and track sections are bolted to each other. The track is obviously a wear item which must be replaced (conclusion based on it being bolted and not welded). The controls are not complicated – some motors for moving the coaster up the first hill, synchronized. Some prox switches to determine where the coaster is. Safety is probably achieved with hardwiring although it would be a good application for PROFIsafe over PROFIBUS and PROFINET.

Ok, time to come clean… the “engineer” was me, a week ago today, in Legoland, looking at the Coastersaurus. [My apologies to those of you waiting for a punch line since this started out sounding like “So a guy walks into a bar and says to the bartender…”]

Then this article was in the local paper: “Lego master modeler at piece with his career.” And of course I loved the pun in the title, but even more, the correlation between Legos and an engineering career:

[Dan Steininger] meets older couples who tell him their children were avid Lego enthusiasts, and Steininger tells them he can guess the occupation of the grown children: engineer.

"Eight or nine times out of 10, I'm right," he said. "There's some connection between what inspires a spark in an engineer's mind, how stuff fits together in space and how a child plays with Lego. It scratches some itch. I see it all the time."

And then Mike A over lunch talked about a recent meeting he presented at where all the major theme park folks had gathered and PROFIsafe was a hot topic.

And finally there was this article in Design News about PROFINET in a theme park: “Disney Rides on Wireless Ethernet.”

Automation Resources

2008-06-20T15:42:00Z

Continuing to answer questions from an Anaheim training class attendee: There are lots of great Internet resources; finding them is the challenge. Of course I’m a little biased towards our site, www.us.profibus.com and the PROFIblog. But beyond those…

The automation industry magazine’s web sites are a good starting point. These are the four I access most frequently:

Automation World

Control Engineering

Control

Industrial Ethernet Book

I’ve also signed up for some of their email newsletters.

Wikipedia is a surprisingly good resource for the more general overviews; for example, PROFINET, Ethernet, fieldbus, and HART.

For keeping up-to-date, this blog and the roughly 30 other industrial automation blogs are good. I keep up-to-date using RSS feeds from these blogs. For more on RSS and an importable list of the RSS feeds I use visit our RSS Primer. Other “real-time” sources include the PROFIBUS and PROFINET forums and the long-running Automation List from www.control.com.

Obviously, your vendors’ web sites will be valuable. Readers are welcome to jump in with additional suggestions.

The knowledge gained on your six-year sojourn into IT should serve you well back in Automation. Industrial Ethernet is being rapidly adopted in industry and that IT knowledge is definitely transferable. The need for cooperation between IT and Control Engineering is one of my pet topics. I’ve blogged about it frequently with some references to magazine articles on the topic:

Control Engineers Boo IT

Control Engineers Boo IT – Continued

IT vs. Control Engineering, again

IT and engineering - two organizations separated by common technology

Balancing the Needs of Engineering and IT

Future FISCO Star Trek IT

Yikes! Yesterday’s post was number 200, a modest milestone, but a milestone none the less.

“And the winner is…”

2008-06-19T21:40:00Z

An attendee from our Anaheim PROFINET one-day training event made a comment to my blogged report which I’m promoting to this post to answer:

“As a former automation engineering in the process industries, and a DeltaV user, I have been pre-programmed to believe that FF was the only way. After a few years in core business IT, I want to get back to my roots in automation and control. I would really like to find out how the buss wars ended (or who is winning) while I had my head in the sand for six years. Any good internet resources you can direct me to?”

Who won the fieldbus wars? Absolutely and unequivocally - PROFIBUS. With 23,300,000 installed nodes through 2007, PROFIBUS has about as many nodes installed as all the other major fieldbuses combined. And the year-over-year growth rate was 24% indicating PROFIBUS growth continues unabated. If you look more narrowly at fieldbuses, just in the intrinsically safe physical layer, that’s where FF plays. And that’s where PROFIBUS PA plays as well. In this area, it’s more of a tie with both in the 700,000 range. (FF reference. PROFIBUS PA reference.) Note that FF and PA use the same physical layer; only the protocols are different from each other. The PROFIBUS PA protocol is the PROFIBUS DP protocol.

I’ll complete the answer tomorrow with Internet links for information in the automation and fieldbus areas.

Offhand Noise Comment

2008-06-17T16:09:00Z

Funny coincidence: yesterday I mentioned using fiber optics for noise immunity (and thinking I probably should have expanded that thought) and today’s email brings reference to an article at Control Design, “More Than a Backbone.” The article meets my deserves-a-mention criteria by presenting some useful information… and using one of my favorite p-words – PROFIBUS. And of course every time it says “Ethernet” I think “PROFINET.” This is a logical substitution; PROFINET uses standard unmodified Ethernet after all. So, for more on using fiber optic cable instead of copper read the article.

Anaheim PROFINET Training Report

2008-06-16T19:58:00Z

It was old-home week for me last week in Anaheim. I lived in Orange County from 1994 until 2000 while working at Wonderware. So after catching up with some other Wonderware alumni on Tuesday, it was off to the training class on Wednesday.

I think we have these PROFINET training sessions down pat, but it’s always interesting to get attendees’ feedback. We added some additional demonstrations to show built-in diagnostic capabilities for the first time. (I already blogged about the feedback that instigated this.)

Some typical feedback from Anaheim:

“Would have liked more process industry focus. Otherwise, excellent seminar!”

“In addition: Comparison against FF – you steered clear of this topic.”

Good news for those seeking more on the process industry and a comparison with FF – there’s a full day PROFIBUS in the process industry class scheduled for Long Beach on September 25.

We had a few folks looking for more details on determinism, motion control, integration with PROFIBUS, integration with GE Fanuc products, and more. We also had more than the usual number who thought the course was “too long.” You can see our difficulty. We try to keep the topics of most general interest in the one-day class, but for real depth we offer the full week certified network engineer classes.

We had several complaints about the cost of parking. For those folks, please remember there was no charge for the course, for breakfast, for lunch, or for the bar afterwards.

On the positive side: “I thought this free training was a good way for our company to learn about the technology and the vendors. Thank you.”

We also had a comment about switches: “One thought when talking about industrial switches – you may want to let people know what features you need for managed versus unmanaged.” We do actually spend some time on that and I’ve blogged on the topic before. To recap, you can use an unmanaged switch with PROFINET, but you’ll get some diagnostic benefits from a managed switch. For a managed switch, we recommend the QoS feature (IEEE 802.1p/q). Useful (while not required) features include trunking, VLAN, and port mirroring. PROFINET does not require IGMP Snooping.

We also had a comment about noise. We talk about the benefits of shielding, but this time we forgot to mention that for real noise immunity you can use fiber instead of copper.

Mike A is in Tulsa tomorrow for the PROFIBUS one-day training event. You can still register for that today if you can be there. My next PROFINET class is August 4 in Scottsdale, followed by a PROFINET Developer Workshop on August 5.

PTO General Assembly Meeting 2008

2008-06-04T21:47:00Z

Automation World’s Process Automation Newsletter today reprised an article from the magazine’s October issue: “Fieldbus Reigns in Process Control.” I liked the article when it first came out; I even blogged about it then.

A quote from the article:

“Plant managers at the water plant in DeKalb Country, Ga., learned about the benefits of fieldbus networks quickly when they installed Profibus in their new water plant. They were so pleased with the results, they retrofitted two existing wastewater plants with the fieldbus. “It makes your life easier,” says Merat Zarreii, F&T Division manager at the DeKalb County Department of Watershed Management in Stone Mountain, Ga.”

Is there a connection between the article, the quote, and the PTO General Assembly Meeting? Yes, indeed. Merat will be presenting an expanded version of his experience with PROFIBUS at the GAM.

Wireless Reader

2008-05-28T21:42:00Z

Trying to make sense of the wireless milieu? Good luck! Here’s my attempt to help with some commentary and links to recent articles on wireless in automation:

First, remember that there is more than one type of “wireless.” I classify three types: RTU, sensor level, and backbone. By RTU, I mean radio connection to Remote Terminal Units typically found in the field like at a water pumping station. It reports status and values via proprietary radios. There is no standardization of the radios although the protocol is frequently Modbus RTU. Backbone defines what PROFINET uses. Because PROFINET is standard IEEE 802.3 wired Ethernet, it is easily connected wirelessly via IEEE802.11 WiFi. This is suitable for bridging difficult-to-wire gaps. Its cost and power requirements make it unsuitable for sensor level wireless. Sensor level wireless has been proprietary in the past but now WirelessHART provides an open standard for this application. ISA is also working on a standard for this space, too. There is some amount of politicking between supporters of the two standards.

Here’s a good article that showcases some of that back and forth between WirelessHART and ISA SP100: “Unplugged: Developing Standards for Wireless Automation” by Maribel Rios in the May issue of Pharmaceutical Technology. After all the point/counter-point comes this conclusion: “The thing to remember is that it doesn't matter who puts out the standard, it's what the users adopt that is going to be important.” [While you’re there, click on the sidebar “Which Wireless” for an expansion of the discussion on the kinds of wireless.]

There is one correction to make about how PI will proceed. *** Caro says in the article that PI “appears to be leaning toward founding the wireless version of Profibus on top of the WirelessHART protocol.” We are not going to create a wireless version of PROFIBUS at all. Instead, PI has decided to collaborate with Fieldbus Foundation and HART Communication Foundation in a unified approach to wireless. (See the original announcement from last September and the recent update. I also blogged about this recently.) We are working with FF and HCF to create gateways to FF, PROFIBUS, and PROFINET from WirelessHART. Once the spec for ISA SP100 is done, the collaboration team will determine how to approach it.

For an ISA-oriented view of the situation read the Control magazine article “Which Way Wireless?” by *** Caro. For a strong users’ advocacy perspective, follow Walt Boyes’ blog – he can’t go long without raising the topic. PI has also come down on the side of the users by foregoing a PROFIBUS process wireless sensor network of its own and instead collaborating with FF and HCF.

A more generic overview of wireless is in the Industrial Ethernet Book: “Industrial Wireless: Unplugged Version of the Factory Network.”

For those of you concerned with the security of wireless networks, read “Wireless Security on the Plant Floor”, also in the Industrial Ethernet Book. This article is by ProSoft’s Wally Gastreich.

Come to the PTO General Assembly Meeting in August in Scottsdale for more discussion on these topics including presentations from Wally Gastreich on wireless security, from Ron Helson of HCF on the wireless collaboration effort, and from Tom Burke of OPC Foundation on the EDDL Cooperation Team (ECT). Collaboration and complementary technologies is one of our themes at the GAM this year.

Raleigh Report: Needed – a Diagnostic Tool for Idiots

2008-05-23T16:49:00Z

Isn’t it just always the case? You brag about something and it completely turns around. Phoenix temperature was 110 when I blogged on Monday; now on Friday we may get all the way up to 70… and rain – very unusual for this late in May. Ok, back to PROFINET. The final point I wanted to respond to from a Raleigh course evaluation form, a very-well presented something new:

“PROFINET (and any fieldbus) communication system needs a diagnostic tool for idiots. A simple, generic plug-it-in-here hand held terminal that will tell a third shift mechanic where the problem is. This is a person that does not have access to a laptop and “programming” software. This person at most has a DVM. The troubleshooting tools you show are aimed at an engineer doing the troubleshooting. I do not want to have to go into the plant at 3am to troubleshoot the problem with a laptop.”

We do show tools that are intended for use by an engineer. But some of the things we showed can help that third shift mechanic without requiring any other devices! This is something we were obviously not explicit enough about. We showed Wireshark, a tool for examining Ethernet frames. This is obviously not a tool for the mechanic. In fact, it’s unlikely to be needed by the user engineer unless the problem is really, really esoteric. It is very much needed by the developer of control devices with a PROFINET interface though.

There are several “tools” we talked about that can be used to provide information the third shift mechanic needs and that information can be presented on the HMI – no plug-in hand held needed! There are two enablers for this capability; they are really not separate “tools”: SNMP from the IT world and PROFINET IO diagnostics from the automation world. First though we need to point out where the potential network problems are. Mostly they’ll be device failures or network problems caused by cabling problems, etc.

Device failures can be diagnosed in the controller and displayed on the HMI. For example, a module within a remote IO rack fails. We demonstrated how this can trigger an alarm during the class. [In future classes, we’ll add showing this in an HMI to make clear this point.] What if the device that fails is an Ethernet switch? I recommend using managed switches with PROFINET IO capability built in. (You can use an unmanaged switch without PROFINET IO capability, but finding a problem is going to be much more time consuming.) You can use the same PROFINET IO diagnostics to find a problem with a switch and react to it in the controller and display it on the HMI. Switches with PROFINET IO capability are available from several manufacturers.

Diagnosing network trouble can be pursued one of two ways: SNMP or PROFINET IO. PROFINET IO functionality can be handled the same way as for device failures. SNMP can be used in the IT way independently of the control system - just Google “SNMP Tools”. Or, better, bring the information available via SNMP into the HMI using OPC. One way to do that is to use an SNMP OPC Server like the one from PTO-member Kepware. Another approach would be to use a separate software program from Network Vision: IntraVUE.

For further reading on SNMP, I recommend the article “Engineer or Network Administrator?” in Automation World.

I blogged previously on other tools, but I know there are others out there. If you know of one, please comment here.

OPC is NOT dead

2008-05-19T18:01:00Z

The plane landed too soon Friday and cut short my OPC story. So, here’s the conclusion:

A number of people have raised the question about OPC UA replacing PROFINET CBA (Component Based Automation). First, a reminder of what CBA is – an easy-to-use, peer-to-peer integration technique. It solves the problem of integrating best-in-class machines from different vendors (with different control equipment on them) into a single production line. This is a unique feature of PROFINET that allows peer-to-peer integration without programming, among controllers from different manufacturers, even using different fieldbuses. Some controllers have this capability built in; those that don’t can still connect using proxies. A proxy is similar to a gateway, but rather than being the product of a single vendor, it’s defined in the PROFINET spec.

To confirm my conclusion that OPC UA is not a replacement for CBA, last month in Hanover I had a conversation with Tom Burke, OPC President. OPC UA definitely provides the ability to integrate diverse controllers when integrated into those controllers. But asking it to work in 10 milliseconds is not reasonable. (I’ve used 10 milliseconds as an example here; PROFINET CBA can certainly work faster – one millisecond is not unusual.) OPC UA does not have the proxy concept for using controllers that do not natively support it. PROFINET CBA does and, in addition, it provides a framework for configuring interconnections in a graphical environment. Downloading those interconnections does not even require taking the PLC out of run mode. OPC is still in the picture with PROFINET though; OPC the only logical way to move data into a computer, be it for HMI or historian or whatever purpose.

Coordinating equipment within one production line: how do users make those interconnections today? This was the topic of an interesting side conversation in Raleigh. The one word answer: wire! Connect a wire from a digital output on vendor A’s PLC to a digital input on vendor B’s PLC. And probably not just one wire, but a bunch of wires. This is fine for binary information, but what about more complex data like tracking information? Now you have to use vendor A’s PLC programming software to create a means of communicating with vendor B’s PLC. Of course, you have to use vendor B’s PLC programming package to create the receiving end of that information. Now you can use PROFINET CBA instead. That saves a lot of time, is easier to maintain, and is more flexible when the inevitable changes occur.

I still have one more interesting topic from Raleigh to cover. I’ll get to that in a day or two. In the meantime, I’ll enjoy our early summer temperatures here in Arizona. We hadn’t broken 100 degrees since last September 21… until yesterday. Today’s temperature should break a record, topping out at a predicted 110. (But it’s a dry heat.)

Raleigh Report: OPC is dead

2008-05-16T21:16:00Z

As many of these PROFINET one-day training events as I’ve done, there is always something new that arises. There are also some things that are always the same. We always get requests on the course evaluation forms to do some hands-on training. With an average attendance of 80 or so, we just cannot carry enough equipment (or people) to fulfill that. We always get contradictory suggestions, too – some want more Ethernet Basics, some want less. This is because some attendees have previously viewed the archived Ethernet Basics webinars… and some haven’t. In Raleigh, many had. We will make the Ethernet Basics review shorter in the future though. So if you are coming to a PROFINET one-day training event in the future, please view the first three webinars before coming. We always get some praise and some constructive criticism. Both are appreciated. Let me share some of each with you:

“Great job! I’ve heard it was good – I agree.” [Funny, I overheard a couple people talking before the class saying something very similar.]

“Good job. Kept it interesting. Excellent price = free!”

“Well worth the time.”

There were a number of suggestions on changing the order and/or time spent on specific areas. We take these into consideration as we update the course material.

On the negative side:

“Safety – A lot was skipped over in favor of opening the bar! NOT a good first impression.” I did the safety presentation and I did skip several slides that I thought I had already covered the points of. Safety could fill a whole day by itself, so we only provide a very high-level overview. We didn’t really rush it to get the bar open; we did rush it a bit to finish at the scheduled time of 4:30. I did describe it as rushing it in favor of opening the bar but that’s just my alleged sense of humor – I don’t drink. Oh and yes, we do have a bar. We encourage attendees to stay and talk to us and the exhibitors after the formal training is over.

On the something new side that demands an answer:

“Worth the time – OPC is dead!!” Whoa, I don’t know how we left that impression – the OPC part, that is. OPC and PROFINET are complementary. We do present an application story where the number of OPC Servers was reduced because of PROFINET. Perhaps this person thought that PROFINET CBA (Component Based Automation) would replace OPC, but that’s not the case. OPC was not designed to move data at a speed of 10 milliseconds. PROFINET CBA is. PROFINET CBA can easily integrate diverse controllers in a peer-to-peer architecture. That way just one OPC server is needed, but it is still needed!

There was one other really good comment that needs some explanation, but the plane is about to land, so it will have to wait for my next post.

On the Road to…

2008-05-14T15:22:00Z

Another airport today – Phoenix Sky Harbor – waiting for my outbound flight to Raleigh for the PROFINET one-day training event tomorrow. If you’re in North Carolina, I hope to see you there. Mike A was in Richmond yesterday for a PROFIBUS one-day training class. It was well-attended with fewer no-shows than our secret formula predicted. I hope the same applies in Raleigh.

I heard from Igor Jovanovic of Omron with some clarifications to my report from Hanover Fair. You can revisit the original post for those updates.

Some Scary Numbers

2008-05-09T21:53:00Z

A recent poll by Control magazine shows that only 12% of users report using primarily digital buses. Keith Larson mentioned the report in the article, “Industrial Networking: The Next 20 Years”: “…a just-completed survey of Control readers indicated that a full 54% of survey respondents still use mostly 4-20mA analog hardwiring for instrument communications. Only 12% reported using primarily digital buses.”

I wish I knew if they were talking about their existing infrastructure or new projects. If new projects, then I am really baffled. Fieldbuses have been around long enough (PROFIBUS started over 20 years ago) and are widespread enough (over 24,000,000 PROFIBUS nodes installed so far) that this should be an easy decision. Is it that there are too many choices? Well, in North America there are really only two choices in each category:

Discrete: PROFIBUS DP or DeviceNet

Process: PROFIBUS PA or Foundation Fieldbus

Industrial Ethernet: PROFINET or Ethernet/IP

PROFIBUS DP and PROFIBUS PA have different physical layers, but use the same protocol. PROFINET easily integrates with PROFIBUS and other networks, too, including DeviceNet and FF. To me the choice of all PROFI looks easy, but I may be ever so slightly biased. So don’t take my word for it – try both choices. Pick a couple small projects, educate yourself about both alternatives, then implement both. This requires some time commitment, but it’s better than staying stuck in the past. I’m pretty confident which technology will win the comparison.

“If you do what you’ve always done, you’ll get what you’ve always gotten.” If your facility will not benefit from easier installation, faster commissioning, quicker troubleshooting, and more uptime then by all means keep use hard wiring. If it will, then get on a bus!

Looking at very different scary numbers, there was a really interesting/scary article in Computer magazine: “The $100,000 Keying Error” By Kai A. Olsen. It’s all about what can happen when your manually entered numbers are un-checked. Think HMI instead of banking and it applies to our automation world, too.

From ABB Automation World, PROFIBUS and PROFINET

2008-05-01T23:50:00Z

Sitting in another airport (Houston Bush) waiting for a plane delayed by weather, here’s my report from ABB Automation World:

Many of the industrial automation bloggers were here in Houston and they blogged about the conference. You can find posts from Greg Hale, Walt Boyes, and Gary Mintchell. Plus news items from others. They provide a good overview of the overall conference, but my interest is much narrower. So, I’ll focus here on the PROFIBUS and PROFINET news from ABB.

PROFIBUS is prominently featured in ABB’s architecture. I attended several of the workshops presented by ABB personnel. They give equal opportunity to PROFIBUS PA and Foundation Fieldbus for process instruments, working with the customer to help them determine the best choice. In the comparison workshop they gave what I felt was a fair comparison between HART, FF, and PROFIBUS. Their recommendations follow the lines you would expect: when the user wants “control in the field” they’ll go FF; when they have a large amount of discrete IO they’ll go PROFIBUS. For discrete IO, PROFIBUS DP is clearly their first choice. ABB goes to great lengths to certify third party devices to work with their systems. In the process they provide a lot of value-add like documentation, GSD files, and more. Over 350 PROFIBUS devices are thus provided for, more than for FF or HART.

Speaking of HART, I sat in on Ed Ladd’s workshop on WirelessHART. Recall that PI and FF are cooperating with HART Communication Foundation (HCF) on a common gateway approach to wireless process instruments. Score one for the users. To think we almost did our own wireless process instrument spec. Be thankful for the “almost.” Speaking of wireless spec proliferation, I ran into Walt Boyes and he shared some of the tribulations with the ISA 100 effort. Watch his blog for the latest on that topic.

ABB has an integration initiative that highlights the power of PROFINET. (Well, it highlighted PROFINET to me, anyway.) The initiative creates a common Industrial Ethernet architecture ranging from electrical substation equipment to process instruments with drives and discrete IO in between. The Industrial Ethernet architecture includes FF HSE, a substation protocol, and PROFINET. All coexist on a single network. (That’s the beauty of Ethernet – on the wire many protocols can coexist – think web browsers, file downloads, and emails as a simplistic example.) PROFINET covered the largest application area and its ability to use proxies to integrate other networks was relied upon. ABB has already introduced drives with PROFINET IO connectivity, too.

I rate a conference based on the number of times they say one of my favorite “P” words (PROFIBUS or PROFINET). This conference rates highly – even the WirelessHART presentation added to the count.

Speaking of the word PROFINET, Phoenix Contact did some video reporting from Hanover Fair. They used the right word, too.

Hanover Fair 2008: Hooked on a Feeling

2008-04-28T19:31:00Z

The final numbers for Hanover Fair are just in – in time for this, my final Hanover Fair post:

5,100 exhibitors from 60 countries

200,000 visitors

3,200,000 leads generated (a 15% – 20% increase)

This blogging software won’t let me post videos, so click here to get the sights, sounds, and feelings of the fair.

Speaking of feelings, yesterday we tied a record low in Phoenix. The humidity dropped to 2% - when we say it’s a dry heat, we really mean it. I know not to expect low humidity in Houston! If you’ll be at the ABB Automation World, look me up.

Hanover Fair 2008: PROFIsafe

2008-04-28T16:03:00Z

Now I’m at the airport with time for my next to the last observation from Hanover: I spent some time with PROFIsafe expert Wolfgang Stripf in front of the PROFIsafe display. Here’s the display:

It depicts an automotive example but applies to any dynamically changing process. The demo highlights safety functions in drives, dynamically reconfiguring a laser scanner, and integrating AS-I emergency stop buttons. Incidentally, the red square is a new type of camera-based “light curtain.” If you look closely at it you can see a little guy in a hard hat sitting on top. When you place him on the moving car, the laser scanner recognizes that more than the car is passing through and executes a safety stop. (The laser scanner is the yellow and black box at the top of the panel.) The new drive safety functions show the different options for creating a safe situation. It used to be that your only choice was to remove power, but now you also have the ability to go to a safe speed, a safe position, and more. The emergency stop buttons across the bottom are on the AS-I bus but are brought into the safety system. It’s hard to see in the photo but some of the devices are PROFIBUS devices and some are PROFINET.