|
|
Applications -
Automation Business -
Communications -
Engineering Practices -
HMI -
Information Resources -
LinuxPLC Project -
Motion Control -
Open Control -
PCs in Automation -
Plant Networking -
PLCs -
Programming Languages -
Process Control -
Sensors -
Software
Re: New forum topic - Open Control
Feb 11, 2002 11:14 am, by Bob Pawley
Text :
Curt:
My comments were made with industrial control as the focus. However, IBM, imbedded, and specialty applications such as Pixar, are providing the
structure I was mentioning. Some of this structure and organization is part of the application such as workstations, animation and devices with Linux imbedded.
Within the control world Linux open source has the potential of one plant, within a group of plants, being programmed quite differently than the other plants within the same organization. If new personnel take over the operation of that different plant they would have little or no expectation of what to expect. Similarly, most companies strive to have common practices and procedures across the organization such that moving from plant to plant people know what to expect. This common practice, the definition of proprietary software, is quite handy when solutions in one plant are to be instituted in other plants - or when trouble-shooters move from plant to plant. Linux without some type of common structure does not assure that this level of common practice would survive. Therefore, Linux, less expensive in the short term, could become quite costly in the longer term.
My point is, we do pay a rather hefty price to ensure an across the board standard practice. This price is forcing changes upon the industry.
Bob Pawley
www.automating-automation.com
250-493-6146
Reply
Re: New forum topic - Open Control
Feb 11, 2002 2:19 pm, by Michael Griffin
On February 10, 2002 02:30 pm, Bob Pawley wrote:
<clip>
> Within the control world Linux open source has the potential of one plant,
> within a group of plants, being programmed quite differently than the other
> plants within the same organization. If new personnel take over the
> operation of that different plant they would have little or no expectation
> of what to expect. Similarly, most companies strive to have common
> practices and procedures across the organization such that moving from
> plant to plant people know what to expect. This common practice, the
> definition of proprietary software, is quite handy when solutions in one
> plant are to be instituted in other plants - or when trouble-shooters move
> from plant to plant. Linux without some type of common structure does not
> assure that this level of common practice would survive. Therefore, Linux,
> less expensive in the short term, could become quite costly in the longer
> term.
<clip>
I'm not sure I really understand your point. Linux is just an operating system. If you are going to construct a soft logic system, you also need hardware and application software (the actual soft logic software which you program). The hardware and soft logic application software should be an even bigger concern for your maintenance personnel.
If your company's equipment standards specify a particular set of hardware and software, how does this result in different software and hardware between different plants just because it incorporates Linux? Either the various plants are following your equipment standards, or they are not. If they are not following your equipment standards, it doesn't matter whether the systems are proprietary or open - they are going to be different. If this is the case, I'm afraid you are going to find a bigger difference between different brands of proprietary systems than you will between different Linux distributions.
************************
Michael Griffin
London, Ont. Canada
************************
Reply
Re: New forum topic - Open Control
Feb 11, 2002 11:17 am, by Curt Wuollet
Hi Bob
> Curt:
>
> My comments were made with industrial control as the focus. However, IBM,
> imbedded, and specialty applications such as Pixar, are providing the
> structure I was mentioning. Some of this structure and organization is part
> of the application such as workstations, animation and devices with Linux
> imbedded.
>
> Within the control world Linux open source has the potential of one plant,
> within a group of plants, being programmed quite differently than the other
> plants within the same organization.
Actually we intend to look as much like what exists as possible. While we can't for example use the same setup tools, editors, etc. we are
going to support ladder logic, ST, and possibly the next great thing. In this manner it shouldn't be any worse than AB in one building and GE in the rest. If it were legally possible we could come closer than that. I would hope we wouldn't be held to greater standardization than
the status quo, after all, it's they who are preventing it.
> If new personnel take over the
> operation of that different plant they would have little or no expectation
> of what to expect. Similarly, most companies strive to have common practices
> and procedures across the organization such that moving from plant to plant
> people know what to expect. This common practice, the definition of
> proprietary software, is quite handy when solutions in one plant are to be
> instituted in other plants - or when trouble-shooters move from plant to
> plant. Linux without some type of common structure does not assure that this
> level of common practice would survive. Therefore, Linux, less expensive in
> the short term, could become quite costly in the longer term.
Again, you could provide that common structure in the only way possible now, instead of an all AB shop or all GE shop, you could have an all LPLC shop. Or do you see more consistancy between vendors than I do? Or are you really thinking about the advantages of an entrenched position in the market, large installed base, etc.? If the latter is the criteria, there is little hope for change of any kind. Instead I see the structure migrating a step closer to the consumer with the user community providing a level of support and assurance much higher than any corporation can in these troubled times. Witness Honeywell, TI, the users support each other long after the corporation has flown. Try to at least consider that there are other models that would work and might even work better. Communities don't quit or get merged or go out of business. And they don't have to be profitable in the same sense which makes them more likely to help rather than having you jump through hoops..
> My point is, we do pay a rather hefty price to ensure an across the board
> standard practice. This price is forcing changes upon the industry.
There goes that structure :^)
Regards
cww
Reply
|
