13.11.07 - Third Generation Control Technology
VIPA introduces new control systems with SPEED7 chips for PLC/IPC/Drives - Interview with Wolfgang Seel, Managing Director of VIPA GmbH.
Everyone is talking about industrial Ethernet and the phenomenal performance gains to be achieved with it. What has been somewhat overshadowed, however, is the fact that control technology has also seen a considerable surge in performance in recent years.
SPS Magazin spoke to Wolfgang Seel, Managing Director of VIPA GmbH, Herzogenaurach, about this and other control technology developments.
SPS Magazin: How has control technology developed over the last ten to twenty years?
Seel: I think it is the industrial PC which has had the most significant influence on control technology. This can be seen fairly clearly if you break control technology down into its performance groups. Whereas the industrial PC has made a particular impact on the upper performance field, the control technology market no longer exhibits any growth in the upper market segment, neither unit growth nor turnover growth, despite continuing to show overall growth. Opinions vary, but let´s say of about 6%. In the upper market segment there is a slight degression, which is clearly attributable to the industrial PC. The industrial PC has made an impact on process engineering in particular (with all of its procedural processes) and on drive engineering to a certain extent as well, where a great deal of calculation and visualization is required. Comparing market share figures for the industrial PC with those of the classic PLC, it is clear that the industrial PC dominates and will continue to grow.
SPS Magazin: Does that not give you grounds for concern with regard to your own business?
Seel: Overall, I harbour no concerns about classic hard PLC-based control technology - which is the segment we cover - being taken over by IPC´s over the next ten to twenty years. Although IPC´s are becoming smaller, the fact remains that the lower and middle PLC market segments continue to display major growth. And I believe that this will remain the case in the long term.
In my estimation, this is based on the very simple fact that those servicing and maintaining them do not belong to the PC generation. They originate from the PC era but, due to their training, that are not so PC-biased. This means that they have a more straightforward training level such as a classic technicians training for example, and are able to cope with an PLC much better than with a PC. In its hardware structure the PC, on the other hand, is - I don´t want to say instable, but it lacks longevity. IPC components require replacement three to four times more frequently than hardware PLC components.
The industrial PC will of course continue to become more widespread - IPC growth is, I think, at least twice that of hard PLC´s - but it is the output volumes we need to look at. Five percent of automation technology today is PC-based, with 95% still PLC-based. In terms of units, that 5% will easily be able to grow twice as quickly as the other 95%. It is also a fact that we have already reached the critical mass. We are no longer talking about an emerging market segment, but about an established market segment with current potential. PLC volume alone is said to be around 8 billion worldwide.
SPS Magazin: Does the hardware play a role at all any more? Rather, is it not the case that the core competence of control-systems manufacturers lies in software?
Seel: I would not refer to it as core competence. I am of the opinion that the software currently has an investment potential which needs to be exploited, of course. And that is precisely our position. We ourselves (as non-software manufacturers) look at where our customers have their investment potential and how they can both protect it as far as possible and reuse it.
Our customers are classic STEP7 users who have been programming STEP7 in medium to large-sized enterprises for years, or even decades. This has resulted in volumes which you can´t simply disregard when it comes to converting to an IPC platform today. The move from STEP7 to something else - whether to another hard PLC manufacturer or to an IPC route is unimportant - involves such a high investment and it results in instability of the entire pre-existing programmes.
The amendments required for a programme to run well and error-free often take a long time. This is the stumbling block many are facing. Therefore, they stick to their software, which of course implies that they also stick to their platform, operating system and hardware.
SPS Magazin: What does the switch to an industrial PC have going for it then?
Seel: It is of course relatively simple to switch from Windows CE to Windows XP or to Desktop Windows if you are already in the IPC-world running a standard operating system. On the other hand, if the STEP7 world is your home, what are you going to switch to from there?
There is only STEP7 to switch to, from one CPU to another CPU. We also know from our sales activities that there are customers who would very much like to change to another system and who see advantages in doing so. But for many companies, the whole software conversion simply makes it a no-go. It simply isn´t on.
SPS Magazin: And what are the advantages of PLC?
Seel: Let me answer it this way: Nowadays, we are looking at the third generation of control technology. The first generation was a central PLC with central I/O. It was from here that all the cables went out to all the sensors and all the actuators.
We have seen the second generation over the last decade, or rather the last ten years: a central PLC and a field bus behind it. The field bus extended throughout the whole plant and in principle had I/Os at island stations with a minimum of wiring. This information was gathered and brought to the PLC where it was processed. At that time, communication performance was on a par with the demands placed on the machines.
Today, we have taken the next step and decentralised the intelligence as well. You no longer have a PLC which completes all of the tasks and then aligns the I/O map accordingly. Instead, the tasks in the central PLC are reduced and from there they are transferred to intelligent nodes which are assigned very specific tasks, e.g. an entire wiring portal. That is an island function, something goes in, something comes out, that is what completes an intelligent node. The central control system only wants to hear from the nodes when something is not working. In principle, that´s how today´s modern automation concepts look.
The same also applies to small, high performing PLC´s with a high degree of communication performance. The decentralisation, i.e. the intelligence on site, means that you do not need the greatest processing capacity. However, you do need a high degree of communication performance. When something happens, all of the modules need to be able to interact as quickly as possible in order to be able to exchange the right information. Of course, the basic performance capacity which allows the PLC to initiate and process the communication is necessary.
On the other hand, of course, it is of no use if the PLC is very fast and the information from the bus that an error has occurred comes 10 milliseconds later. By then the PLC is already 20 or 30 cycles further along. This modern control architecture is very well suited to the PLC and this is where it has significant advantages over the industrial PC.
SPS Magazin: What role does performance play and what are the factors influencing it in your opinion?
Seel: I think the performance of PLC´s is basically influenced by two things. The first influence is of course the communication performance deriving from the PC world. In the early days, we had the normal serial interfaces with 9,600 Kbaud with which everybody is still familiar today.
Then 38 Kbaud was the limit at some point, and then there was someone who had an interface going up to 115, which in turn was replaced by Ethernet with 10 Mbaud. 10 Mbaud was the standard for more than ten years. From 1980 to 1990 that was the exhilarating speed with which Ethernet communication operated. And automation technology continually adapted to this speed platform.
A Simatic S5 was programmed with a 9,600 Kbaud serial interface. Then came the first networks which were able to change data with 9,600 Kbaud. The transfer rates were gradually increased in order to achieve higher throughput. However, this was always driven to some extent by what was available on the commercial market in the way of products, chips, technology.
In general it can be said that the performance gains in the STEP7 PLC segment in particular have increased by about tenfold in the last four years. We continually carry out benchmark tests to see where a CPU 318 stands, with which we measure ourselves contemporaneously against technological standards at a given point in time. What has followed in the way of normal microcontroller-based PLC´s, what has followed in the way of Asic-based versions, what has followed in the way of high-performing PC/IPC-based solutions?
It can be said that, in the stated period of three to four years, performance has improved by a factor of ten, communication performance included, of course. The Ethernet system with 100 Mbaud then appeared in 2000, meaning that, here too, the gross data transfer rate alone had increased tenfold.
And this means that such conveniences as live-image transfer or similar, are easily achieved via Ethernet, which were not yet an option at 10 Mbaud. Image transfer at 10 Mbaud was always very borderline; the format had to be reduced, the colour depths had to be reduced, etc, in order to be able to transfer more or less reasonable images and these are of course the demands which are now made of today´s automation technology.
SPS Magazin: What developments are you currently seeing in control technology?
Seel: One of the most important influences (also in the PLC field) is processor technology, with the semi-conductor-based speed gains. Almost ten years ago now that was the key reason for us to develop our own PLC processor. The goal was to be significantly faster than the competition, but it is also important technologically.
This technological leap has achieved two things: firstly, modern, energy-saving processor technology results in less heating. Secondly, this technology is extremely fast. In terms of the number of cycles alone we are in fact relatively slow. Our processors work at a frequency of 48MHz and/or 96MHz and use a maximum of 2W. This means that we are also able to cater for extended temperature requirements without any problem.
One of our objectives was to take PLC performance so far that a PLC command could be carried out with an internal clock signal. This means that, if you are talking about a processor frequency of 48MHz, a command can be carried out 48 million times per second. Our competitors´ products need ten or twenty internal clock signals for one PLC command, for example. That is the huge advantage we have achieved with this PLC on chip development.
SPS Magazin: New PLC processors from Profichip for the PLC/IPC/Drives are currently being developed. These are to be named 7100 and 7200. Can you tell us something more about them?
Seel: Exactly, the new generation is called 7100 and 7200. The two basis processors will each be available in different versions. First of all, they contain a PLC core, as is already the case with the 7001 now. The chips also contain digital inputs and outputs, technology functions with counters, multiphase motor functions, pulse width modulation, etc, with analogue instrumentation and drive technology also being available.
All of that is already available and will also be included in these new modules. We have more PROFIBUS masters on board with the new chips, we have CAN interfaces on board, we have Ethernet interfaces on board for real-time Ethernet. On top of that we have a whole series of standard interfaces such as UART or I2C on board, which enable more straightforward communication tasks. And of course we have integrated a peripheral interface with which external peripherals can also be connected.
SPS Magazin: Will the chip also contain MPI?
Seel: MPI is a PROFIBUS subset. We have three concrete PROFIBUS masters on board and a PROFIBUS slave, which is integrated directly into the hardware, derived from today´s existing PROFIBUS slave modules. These are referred to as a core, meaning that, as has been certified by the PROFIBUS USER ORGANIZATION, the PROFIBUS slave, the VPC3+C, is contained in this processor.
This means that all of the functions that a slave module can perform can also be performed by this processor. Using it solely as a slave system would of course be relatively expensive because the module contains the entire PLC memory for the small, 7100 solutions. The module already has 256KB of PLC memory internally so that it can meet the performance requirements of a micro-PLC through to a medium PLC - if we are looking at the Siemens raster, through to the CPU 315.
SPS Magazin: Where is the journey headed at VIPA and Profichip?
Seel: VIPA´s competency is without doubt the construction of control systems. With the availability of the 7100 and 7200 chips, in the STEP7 CPU segment we want to convert all existing VIPA 100V, 200V and 300V systems to SPEED7 CPU´s entirely.
Under the 100V System there is to be a new product family, namely the Micro System. On the one hand the Micro System encompasses I/O peripherals which are also equipped with today´s modern channels, i.e. Ethernet-based communication. The Micro PLC is integrated into the decentralised peripheral. A PLC will also be integrated into a node housing corresponding to the present-day performance capability of an upper S7-300.
This product will be able to meet market demands for slice I/O´s - as we call them - which can be very finely tuned. We have set ourselves the goal of introducing the PLC/IPC/Drives system in 2008, for instance, and that it should be ready to go into production six months later. The chip itself will be detached from VIPA interests and sold as an independent product by Profichip, as a "PLC for everyone" which can be used for any kind of individual electronics.
When you look at the situation today, there are numerous embedded platforms which have the major disadvantage of either needing to be programmed by an engineer or needing a very complex soft PLC installation. This is relatively complicated. We will supply the chips with a software package which enables the customer to programme his application like a PLC.
This means that the user can simply programme his application with STEP7. This also has the further advantage that even a service provider´s customer can have independent access to the programme if required. It is far more likely, however, that he will have someone who can programme STEP7 than someone who is an expert in C++ or the like. Wherever on-board electronics are required, we want to be able to put the tools in peoples´ hands to be able to create on-board electronics independently with relatively simple PLC technology.
This product will be distributed worldwide. I think it is going to be one of the very important milestones in establishing STEP7 as a global standard in its current form. (kbn)
The interview was carried out at VIPA in Herzogenaurach by Kai Binder in July 2007.
- Links:
