In defense of Big Brother

P

Thread Starter

Philip Wilkie

Its time for a break from my other job, so having trolled through the archives I thought to write a little burley bait and see if I catch any flame fish by coming out in defense of the "big companies" and their evil schemes for world dominance.

There are three main types of people who play roles within any company, the "managers: who endeavour to set the agenda and regulate the money flow, the technicians who get things done, and the entrepenuers who woo and sell the customers. I willing to bet that most of us on this list are "technician" types who love getting things done. The centre of our world are the toys and tools that we are skilled at using, and we take pride and pleasure in a "job well done".

But it means also that typically we utterly fail to "grok" those other creatures that infest our lives: the bean counters, the project managers, consultants (hoick-sput!), the sales jocks and marketing dudes. And so many of the opinions I've read in here, while usually sound technically, or valid within a limited sphere, completely fail to make it in the real world because of this.

The fact is that the Rockwells, Seimens and Schneiders that dominate our industry do so for one simple reason...that 85% of the time they deliver to 85% of the market, 100% of what it needs: predictable, reliable products with stable life-cycles, that are used in a very wide range of applications, all at a reasonably low total cost of ownership. i.e. good value. Within those parameters they have been pretty successful. Not perfect, not leading edge, not always elegant and not always the lowest purchase cost, but in the final analysis (measured in the bottom line of customers who use them)... damm good value.

Let me take one example that the archives throw up...the question...why can't I connect the Ethernet port of my SLC to the same on my Quantum or S7 and get data to shift? Perfectly good question, but the answers are of often muddied by silly prejudices about how monopolies and evil marketing hype are used to protect "weak products" and milk cash from captured customers. Sighs...if only the world were so simple.

Here is a simple question. When Modbus, or DH+ were first introduced 15-20 years ago, how many qualified networks engineers were to be found in your typical electrical crew? Not so many right? The fact is that millions of feet of "Blue Hose" and Modbus was pulled in and connected with reletively little fuss (except when the basic rules were broken) by guys like you and me who would not have known back then what an IP address was if it bit us in the butt. The upside was..it worked; and if it didn't... you knew whose cage to rattle.

The downside was that the "big boys" didn't like getting their toys broke, so for a long time, too long, they kept them locked up in their proprietary corporate boxes. Well that was way back then...times have moved on. Overall the industry is now capable enough to manage open protocols and hardware systems with far more maturity than 20 years ago. (The PLC was only invented 30 years ago.) No-one has been more aware of this than the major vendors, who have vigorously responded with DeviceNet, ASI, Profibus DP, Modbus TCP and a sprinkling of other offerings promoting open interconnectivity.

In response we grizzle about "network wars" blithely assuming that somehow a committee from all the big competing vendors was supposed to sit down and all peacable like and dream up (and in a few short months) some universally wonderful protocol that was a) technically advanced enough to be future proof, b) capable of allowing each vendor to add their own unique brand value (no-one wants to turn their own product into a commodity...when did you last ask for a drop in pay?) and most importantly c) didn't leave behind and render obsolete their own huge installed bases.

What did happen was that over the last decade or so the vendors have all made big strides towards a version of the perfect protocol that they could live with. Siemens evolved various incarnations of Profibus; Schneider re-packaged Modbus with TCP/IP wraps and added sexy Weblike functions, and Rockwell slowly woke up to the fact that the CIP protocol they first invented for the ControlLogix backplane could get used in other forms, such as DeviceNet, ControlNet and now EtherNet/IP. All these major corporate initiatives are OPEN...to some degree or another. The days of the proprietary protocol are gone. When I started in this game (not so long ago:) if someone asked me, can I connect Vendor A's Box to Vendor Kamakuza's ThunderBox B?... the answer was usually no. Nowadays the answer is usually yes. (Well at the very least you might have part with a little cash.)

Now the big vendors haven't been spending shareholders cash on these OPEN Protocol programs, just so as us techies have new toys to play with. Along the way the marketing and sales guys have to build a critical mass of customers to place orders and roll in real cash to pay for it, the managers somehow have to hack out a business plan so as it looks good in Powerpoint and we all have to show the CFO that it makes an ROI. (And besides there's all those corporate AMEX's to pay for). Is this so wrong? Another simple question. You are about to buy a brand new car, one you really want, you've earned it, you got the cash... but the scuttle is... the works that makes them is slowly going bust. Now how do you feel about shelling out those hard-earned greenies?

My guess is that in the next few years the natural action of the marketplace will gradually nudge the three biggies into some kind of accomodation with each other about interconnectivity. Personally I hope that Rockwell's CIP (Control and Information Protocol) is the winner, but that is by no means certain, given the way the world works.

CIP has the following merits. 1. It is fundamentally routable across multiple network segments. 2. The publish/subscribe messaging paradigm uses bandwidth intelligently. 3. It defines an set of open automation objects and services that are the foundation of true interoperabilty. 4. It is media independent. 5. It's not old sauce served up as new soup.

But this is just my little day-dream. Meanwhile back in the real world the issues move on. In a few years time we will hopefully have buried the silly PC v PLC debate, the industry will have a least one set of open services and protocols that most of the big players support, and we will have a univeral logic editor that scans text based FD's and generates logic code and downloads to anyone's hardware. Whoa damm! Those polar bears do fly low.

Guys. The core of my argument is this. Those other dudes, the bean counters and marketing stiffs don't care about our leading edge technical elegance, it just doesn't get on their radar. They don't want our latest version of perfection; it's too expensive. They don't actually want our fancy boxes and software tricks at all. What they really want is for us to deliver minimum internal process risk in order that they can both continue to maximise their external (entrepenurial) market risk and at the simultaneously minimising their financial risk; they want us just to invoice them for a job well done, and not to make trouble. i.e. the age-old balance between risk and return.

And it is a huge marketplace out there. Automation customers come is all sizes and shapes, with all kinds of value drivers; good, bad and ugly. Some of them will love Linux and the OpenPLC, but the remarkable thing is that the "big boy vendors" in our industry keep most of these customer's handle's cranked most of the time...and along the way us techies siphon off our own cut of the tomato sauce. There is nothing wrong with constructive grizzling, informed whingeing and the odd well deserved shot or two(especially if the fan is running), but it's time we lost the immature envy script that goes like, "Bill Gates is rich, I'm poor; therefore Bill is a bastard."

It's a big wide world out there. Have fun... my word these daisies do smell funny.
 
J

James Ingraham

An interesting diatribe. A few points about why I don't like Big Brother, aside from the obvious "love the underdog, hate the oppressor" paradigm you seem to think we all operate under.

Allen-Bradley: Several occasions, A-B has sold me products that would not work for my application. Some A-B people knew this, but the massive size of A-B prevented the right people from knowing the right thing at the right time. In the end, (on several occasions) we ate big chunks of money over this. A quick anecdote; when considering a switch to ControlLogix, I was sitting in a room with 4 A-B people and 2 people from my local A-B vendor. I commented that I was most interested in Sequential Function Charts, which I knew were supported because all the docs say "IEC 61131-3". Of course, the fine print in IEC 61131-3 mentions that all the languages are optional. Naturally, the A-B docs don't point out which languages are or aren't supported. And none of the 4 A-B people in the room mentioned that SFCs were not supported. It was my vendor who called me a half-hour after the meeting to let me know that I wasn't getting SFCs any time soon.

Modicon: There are a lot of things about Modicon that I like. Their software is not among them. Modsoft is unusable since all our systems are NT, 2000, or XP. Concept crashes like crazy. We waited FOUR WEEKS to get their OPC server, which didn't have some of the features we had used in an applicom OPC server. I'm sure that there is someone at Modicon who knows their product line. I have not talked to anyone who does. I've been asking for information about their servos for five years; I have never gotten anyone who knows anything about them.

Siemens: I haven't dealt much with Siemens, so I can't really comment. What I can say is that their Prodibus-DP DB9 connectors have caused us no end of headache.

Fanuc: Let me include both the Fanuc robot guys and GE Fanuc in one group, which I'm sure both groups would hate. Their prices are literally triple A-B's... and I think A-B is over-priced! Again, if there is someone who knows the products, they are locked in an ivory tower hidden from the real world. GE Fanuc has the nerve to say things like "The Power of Open Systems!" while having the most closed, proprietary system on the planet. (Try running a GE Fanuc servo drive from anything besides a GE Fanuc PLC or CNC or a Fanuc robot controller.)

In short, I don't hate the big guys because they're big. I hate them because they jerk me around, don't respond to me, and don't care about my business ("Puny little OEM? Who cares?").


A quick response about CIP:
>2. The publish/subscribe messaging paradigm uses
>bandwidth intelligently.

Sort of. Once it's done, sure. But scheduling a ControlNet network is anything but "intelligent." Ethernet/IP is slightly better... but you are in deep trouble if you are using the same Ethernet card to program the PLC and you tell it to send more data then it can handle; you are now kicked off-line with no way back in to fix the problem. You have to hook up serially. Having said that, I admit it's got some nice features. I nonetheless prefer the simplicity of Modbus/TCP, desipte being old sauce as new soup.

Someone might reasonably read this and say, "well who *DO* you like?" Nobody. No company has ever fulfilled my expectations in the world of Industrial Automation. I prefer used-car dealers to these guys. But what can you do?

-James Ingraham
Sage Automation, Inc.
 
T

Trask, Robert

clip....
Someone might reasonably read this and say, "well who *DO* you like?" Nobody. No company has ever fulfilled my expectations in the world of
Industrial Automation. I prefer used-car dealers to these guys. But what can you do?
clip....

I used to have this attitude. Then a couple of years ago I bumped into Beckhoff. The hardware is modular, flexible, reasonable. The company is
small enough to be very responsive. And the software kicks butt. It ain't easy, but it works.

The hardest thing is talking end users into using it. I hear a lot of concerns (whining) because it is not A-B or Siemens or GE or whatever they've been suckered into using.

I can't recomend it enough.

"www.beckhoff.com":http://www.beckhoff.com

Robert Trask, PE
MindSpan eSciences
Wilmington, NC USA
 
L
James
An interesting unbiased view of the World. I would like however to give you some inroads to "Modicon" Schneider Electric people who do know what they are talking about. I do this in the knowledge that you will get timely and accurate answers. For the OPC stuff, please contact Jonathan Blanchard on 978 975 9021 or "[email protected]":[email protected] For Software, please contact Wolfgang Langer on 978 975 9742 or "[email protected]", mailto:[email protected] For Motion, contact Richard Covarrubia on 978 975 9777 or "[email protected]", mailto:[email protected] I am sure in every case, we would be more than pleased to help you.

Best regards

Lee J Ward
Business Manager - Programming Software
Schneider Electric
Tel: 978 975 9489 Cell: 978 902 6238
Fax: 978 975 9321 email: [email protected]
 
I was just reading some old articles and i had to comment on this.

> Modicon: There are a lot of things about
>Modicon that I like. Their software is not
>among them. Modsoft is unusable since all our
>systems are NT, 2000, or XP. Concept crashes
>like crazy.

hah. My thoughts exactly. this is the very reason why i refuse to upgrade past win98. Actually, they finally forced me to do it with my laptop, so i bought an extra HD and loaded DOS 6.22 on it. Then, to make modsoft connect, i had to use slowdos, cause 500MHz is simply too fast.

But at any rate, when asked why i still use modsoft, i just say 'Concept is ... well a bad concept' I find no redeeming qualities outside of IEC languages being available. and since I'm confortable with LL984, thats not a big plus. Its ridiculous to have to upgrade the flash every time you turn around. But the thing that kills me is that schneider has access to a pretty good windows based programming software in the telemechanic PL7 program. Now, i dont think much of those PLCs, but the software is thousands of times better than concept.

>I'm sure that there is someone at Modicon who
>knows their product line. I have not talked to
>anyone who does.

I dont know about product lines, but they have some of the best tech support people I've ever come across.

And 1 more thing, the original message was a good one. I think I'm going to print it out and read it very once in a while when i start to feel myself lose touch with reality. i probably know a few people who need to read it right now.


-jeff
 
B

Bob Peterson

My guess is the Frogs have a serious case of not wanting to have something made in America competing against Concept so these products are not pushed, and will probably end up being obsoleted someday.

Bob Peterson
 
B

Blunier, Mark

> hah. My thoughts exactly. this is the very reason why i
> refuse to upgrade past win98. Actually, they finally forced
> me to do it with my laptop, so i bought an extra HD and
> loaded DOS 6.22 on it. Then, to make modsoft connect, i had
> to use slowdos, cause 500MHz is simply too fast. 

You are probably using an old version of Modsoft. I had the same problem when I was force to upgrade my 486 to a Pentium when the laptop failed. The later versions of modsoft didn't have this problem. I'm using 2.62.

> But at any rate, when asked why i still use modsoft, i just
> say 'Concept is ... well a bad concept' I find no redeeming
> qualities outside of IEC languages being available. and
> since I'm confortable with LL984, that's not a big plus.

I agree. The new stuff may be prettier, but in my book, pretty doesn't make it better. As much as people complain about ladder, for control systems it still makes a lot of sense.

Mark Blunier
Any opinions expressed in this message are not necessarily those of the company.
 
nope. I'm using the lastest and greatest. i think its 2.62, i suck with numbers. But i was told a lot of times that what i had was the last version they were ever going to put out.

-jeff (necroposting)
 
Top