Control Strategies - The PLC: programmed obsolescence?
Alan Young speaks up for PC based control topologies in this latest
article in our series on general approaches to industrial automation
Born in the late 1970s, the PLC grew out of the desire to use modern
electronics to reduce cost and to provide more flexibility than
conventional relay based control systems. History repeats itself, and the
same industry is today part of a much wider trend to make use of PC
technology to deliver more processing power and flexibility in an
exceptionally small, low cost package. In the 1970s a computer was almost
science fiction to most people, so it was a good idea to use a 'ladder
logic' based programming language which could quickly be understood by
the plant maintenance engineer. Twenty years later, the PLC has become
smaller, faster and cheaper, and the ladder logic program, now given
official status as part of IEC 1131-3, has established itself as the
backbone of industrial automation real time control. It has given good
service to a generation of maintenance engineers who found it easy to
learn and to fault find at a basic level, but who came to appreciate its
limitations, as control software and the demands of manufacturing
industry became more complex.
The careers of those engineers have developed alongside the PLC and they
now probably manage a small, multi-skilled team who have to deal with a
whole range of modern technology, yet they continue to specify a PLC to
manage critical control functions. Will this strategy hold up through the
next 20 years or will the ever increasing need for smaller, less
expensive, easier to maintain and smarter control systems kill the PLC
and herald the next generation of industrial automation control? Today's
technology has generated discussion on three common themes across many
sectors involved in real time control: powerful, low-cost processors;
fieldbus, and operating systems. Increasing use of PC chip sets is
inevitable, open communication standards are well established and
spreading fast, and although the jury is still out on operating systems,
an evolution of Windows NT/CE or Linux might hold the answer. A key
strength of the big PLC manufacturers has been their ability to sell the
'total solution' on the back of their controllers.
Open communications has broken down the proprietary I/O structures,
opened the I/O market to other manufactures and given the user the
freedom to select 'best of breed' I/O hardware for each element of the
application. Ladder logic PLC languages have been stretched by the number
crunching and operator interface requirements of modern systems to the
point where hybrid configurations combining PLC and PC are quite common.
Hardware implementations have ranged from modules with limited PC
functionality that can be plugged into the PLC, to office style PCs
perched on shelves inside the control cubicle. The old view that a PC
will never work on the factory floor has been overturned partially in
the UK but extensively in the USA and other markets. In these markets the
industrial PC application has managed to set itself apart from the
sporadic reliability of office based PC software. PLC supporters are
finding it much more difficult to deny the hardware advantages of the PC
solution particularly in the face of new built-for-purpose industrial
panel mounted PC's with an LCD screen, a fieldbus card and a range of
powerful software solutions.
They take up no more panel space than a 10in or 15in LCD display (much
less than a CRT display) yet deliver processor power to rival the latest
desk based PCs. In the past it has been tempting to press the latest
bargain desktop PC into service as the controller and to live with the
problems of long term reliability, support, maintainability and how to
mount it securely in the control panel. This approach will need to be
rethought in the light of the latest cost effective industrial panel
mounted PCs. By choosing a Panel PC or Panel Workstation the user has
eliminated many concerns about dust, shock, vibration, temperature,
support and maintainability whilst reducing panel space and getting a lot
of processing power for his money. Who would want to risk using a system
designed for an office with the disadvantages of its CRT based display
when the cost saving is so small? Yet reservations about the use of this
new technology remain. The PLC faction clings to device's only
differentiator, - ladder logic programming, with its promise of easy to
understand, easy to maintain and modify programming. How valid is this in
a world dominated by the Internet and graphical user interfaces? Enter
the 'soft PLC' package.
This duplicates the key benefits of the PLC and its ladder logic
programming in a single powerful and inexpensive PC based controller. The
best of both worlds for the user, and there are a wide range of packages
to choose from. Increasingly, OEMs are looking further ahead and are
preparing to migrate all their software to a language such as C or C++.
The benefits of doing this can be dramatic. The software engineer no
longer has to work with a language optimised for maintenance, and can
instead work with first class software development tools. A stack of
ladder logic printout becomes one page of C/C++ code, and from the
employer's point of view, there is a pool of talent experienced in C/C++
programming available to work on the next project. One language can
handle all control software requirements, which simplifies software
maintenance/support and allows the core of the program to become
independent of the hardware platform. But have we forgotten the needs of
the man or woman charged with maintaining the equipment? Not necessarily.
Their enthusiasm for ladder logic comes from a desire to understand the
control logic for faultfinding purposes. In practice the simple ladder
logic instruction set drives the programmer to produce complex structures
that are not at all easy to follow even for an experienced programmer
with the aid of the original developers' documentation. In any case it is
rarely necessary to get to the heart of the program since the great
majority of production interruptions can be traced to input or output
hardware. Powerful fault finding tools to be developed using high level
languages that can significantly speed repairs, highlighting potential
problems before they shut down production, directing maintenance
personnel straight to the cause of the problem, calling up appropriate
spares, logging faults and even paging maintenance staff. Programming a
system in ladder logic was never quite as simple as programming the
'traffic light' exercise on the training course, but converting to a
system programmed in C++ will also need careful consideration and
planning.
Existing programmers need to be re-trained, or C++ programmers educated
in the workings of the machine. Some careful management is necessary if
deadlines are to be met and the finished product is to satisfy the end
user's need for reliability and maintainability. Manufacturers of less
specialised systems, which are specified in considerable detail by the
user, will follow on down the PC path, probably retaining ladder logic in
the shape of a Panel PC running a Soft PLC package. The end user today
still has to be convinced that the final result is equal to, or better
than the old solution, and he will be looking for a large slice of the
cost saving to make the change worth while. Programmable logic
controllers have enjoyed exceptional end user confidence and support, but
they cannot withstand the force of new technology. Early PC based
concepts were easy to fend off, but new thinking on communications,
operating systems and system packaging have bought the end of the mid
range PLC clearly in to view. As the world's calendars reset to 00,
industrial control system design engineers will be making a fresh start
on a blank sheet of paper.
Alan Young is Business Development Manager for the Industrial Automation