HomePage Italiano  English Search
facebook
Integration close at ... hand
Article taken from "Automazione e Strumentazione" - issue 49 - September 2001

Image 1
The magazine cover is dedicated to Automa

Automa is continuing with the development of new solutions aimed at integrating the two major subject areas, management and plant, relating to company management and industrial automation respectively. Even if the dialogue between the various levels still presents some difficulty linked to structural or system reasons, the architectural model for integration is now close to optimisation.

By Aldo Fiocchi

The problem of integration between the world of industrial automation and the "parallel" one of management type, basically regarding production, has now, for some years been the focal point around which all the studies and research of the companies in the sector have been concentrated: the future, in fact, is increasingly concentrated on the concept of dynamic business whereby the flow of company data must be organised in such a way as to be the fruit of a "strong" two-way correspondence between logistics (the central, strategic and irreplaceable element of an industrial activity) and the various functions contributing to that type of activity.
In this context which is, however, still relatively new, Automa, the well-known company from Pedrengo that has been an important reference point for years in the implementation of information systems for industrial automation specifically aimed at supervision and man-machine interface applications is, with viable solutions, finalising the study started a few years ago on the possibility of truly integrating the four fundamental levels that constitute the framework of a company information system. These levels are: control or PBC (PLC-PC Based Control) and running or HMI (Human Machine Interface) on the Plant side, and programming or EPS (Enterprise Production Systems) and planning or ERP (Enterprise Resource Planning) on the Management side, leading to the birth of integration, a process that Product Manager Mario Borali defined as a "bold attempt" at managing an architecture and a problem area that are both articulated and complex.
Mario Borali can now express his overall satisfaction since five of the business units that have always been the company's "warhorses", namely HMI, MONITEL, CAM, EPS and SPECIALS, constitute a fully defined reality from the integration point of view: these five business units, in fact, form a "technological convergence" that can be compared to the combined action of the five fingers of a hand and, not by chance, the symbol of this new development model is, indeed, an open hand.

The evolution of a philosophy
Automa has always been part of the group of suppliers of top level automation engineering as regards special HMI-based and monitoring projects.
One can, for example, mention the structural and environmental monitoring projects developed for Enel Hydro (formerly ISMES) such as the INDACO project, which has been used to implement numerous dam-side checks, monitoring Etna (Poseidon), the Tiber basin, the Tower of Pisa, St. Mark's Basilica and the Brunelleschi Dome, as well as many others in the structural monitoring area. Also as regards monitoring, mention must be made of the ARGO software package for controlling atmospheric emissions and of other monitoring/supervision software packages developed "to order" for a company that produces security systems. All this, over the years, has allowed Automa to acquire considerable experience in sectors that are, technologically, relatively far apart.
But "the world doesn't stand still" as Francesco Guicciardini was already writing in 1400.
In fact, until a few years ago, management of a fire prevention centre required a totally different technology compared to that used, for example, for dam-side control which, in its turn, was completely different to that used for an automation system: each system represented a "niche" solution, including from the technological point of view, and thus a specific culture had been developed. It is only recently that technological development has allowed the software to become the true reference and meeting point.
It is for this very reason that, from 1991 onwards, the increased concentration of the know-how of the individual sectors in the software part, together with the fact that the software was principally based on PC architectures, has ensured that, in the Twenty-first Century, we are witnessing a technological sector convergence that is impressive, to say the least: today it is in fact possible to develop totally integrated systems based on PC architectures.
With these systems, for example, it is possible to carry out remote surveillance with closed circuit TV cameras and, at the same time, manage signals, shutters and/or fans according to a philosophy that Automa has already used (in part) in '95 with the implementation (in partnership with ISMES) of the Great St. Bernard tunnel monitoring system, one of the first examples of a PC-based supervision and control system where the user was the direct witness of the considerable advantages obtainable with multi-sector technological convergence. What could have been considered an exception up to a few years ago is beginning to make progress as a rule, namely in the spirit of making the treatment of data coming from different situations "homogenous". In the EPS level (and not only MES, i.e. Manufacturing Execution Systems) Automa sees the ideal position for implementing the functions of "homogenisation" of the various sources of data and interface between HMI and ERP.
In other words, it is not only the "processing" that must be an integral part of the information system in its totality, but also the set of technological systems sometimes inappropriately called UTILITIES, but which are often linked to production, such as the control of energy, services, accesses and so on. This is why Automa has already anticipated the choice of "technological convergence" by several years, seeking to define an architecture on which software and further skills can be built with this precise objective.
At this point all that remains is to examine the offer of this "open hand" in detail starting from the HMI part.

HMI: small, highly reliable and. . . compact
As regards the HMI part the company from Pedrengo, starting from Super-Flash and the line of products relative to Supervisione Libera, has extended the offer even further focusing its interest on the SoftPLC sector and, in addition, on everything concerning the so-called "Embedded World": in fact the company's wealth of experience acquired over the years is well known as regards the development of applications specifically dedicated to the HMI part so as to build man-machine interfaces that are indeed compact but, above all, viable, namely systems intended to operate on very small, highly reliable PCs, thanks to the implementation of a DOS-based, wide spectrum, 32-bit standards proprietary operating system that is solid and, much more important, with a high performance.
Regarding the "Embedded World", however, Automa has won a first victory by putting into practice an idea based on the use of a real-time version of the Linux operating system to create the first Linux real-time/embedded integration, the heart of a set of hardware and software architectures that can be further extended in the future. Thus Scatolinux was born, an acronym for Scalable To Linux, a clear expression referring to an object able to perform a vast number of applications (multus) as a data collection, validation and recording system, a controller for soft real-time applications, a field Web server and a router with local or remote I/O management (to name but a few). The whole, one could say, being almost concentrated in a dot, since its "physical" form is that of a cube with sides of only 13 cm (. . . in parvus), but, above all, an object of great help to the user in developing his own data processing or interface systems, for the very reason that it is provided with "absolutely" standard hardware and an "absolutely" standard operating system (indeed the system's corner headers are PC104 and Linux standard).
Scatolinux is also highly reliable since it is not fitted with a hard disk, it can be inserted on a guide, it is set up with industrial criteria at PLC level but, at the same time, can also "run" a supervision system such as Super-Flash internally and can be easily configured and programmed using a Web-based application or a network or serial terminal emulator. Thanks to its special internal architecture Scatolinux can be connected to both the process (field bus and building automation, in the latter case as a control centre for managing surveillance, fire prevention and access control systems and so on) and to the outside world (Ethernet and Internet in general): data management also passes through a "history", a specific function able to store data "from the field" independently, even if the higher level is missing. Access is possible at any time through a Web Server (Website).

For history
Automa's interest in Linux arose a few years ago, namely when Linux was considered little more than a curiosity, following a similar interest (together with considerable investments) that the large companies in the sector were beginning to show for this product, even though application of the Linux philosophy was still a long way away from the "Automa world" but with the knowledge, however, that the level of reliability of Linux was very much higher than that of equivalent Windows systems.
Linux unexpectedly leapt into first place in Automa's scale of primary interests when the first Linux real-time projects arose: RTAI, a Milan Polytechnic project and RT Linux (Mexico City University).
But the "intrinsic" feature of Linux which was the "turn of the tide" in Automa's business choices was, in fact, that relative to the possibility of developing embedded applications due to the availability of a microkernel in the systems that calls the services, loading and unloading them in memory: this, together with the subsequent appearance of suitable development tools on the market that act as a support for even the least experienced user, allows the operating system to be "pruned" to make it as compact as possible.
The value added by Automa as regards Scatolinux consists in "fitting in" not only the complete operating system but also a Telnet Server, an FTP Server, the field bus drivers (Canbus, Profibus and so on), drivers for a maximum of 10 serial ports and the TCP/IP drivers. All this "compressed" into only 5 MB of disk memory, a size readily available on disk-on-chip or on flash disks and on small sized PCs. The system is also "ready for use" within about twenty seconds from the start-up of the application and can be "shut down" equally quickly without any limiting shut-down procedures.
Thanks to a partnership with VIPA (Gesellschaft für Visualisierung und Prozessautomatisierung mbh), an important German company specialising in the production of Siemens-compatible systems, Automa's future possibilities as regards embedded applications will multiply: the VIPA System 200V Compact IPC-288L will be used to develop an even smaller "Scatolinux" (50.8 x 76 x 80 mm). With its 32 MB DRAM, 32 MB of disk-on-chip, flash disks, microdrive, Ethernet interface and field buses it can offer the power of a 486 processor in an extremely compact industrial unit.

MONITEL: farseeing eyes
On the basis of what has already been said, now more than ever, Automa considers the prospective development of the monitoring and remote control section to be of great strategic importance because it will be the natural base on which to build a versatile business. All to the advantage of, for example, large services distribution Companies of the calibre of ENEL, Telecom and so on. As regards this, there are concrete examples such as BAS of Bergamo which, already having a structure of fibre optic cabling distributed throughout the city serving their own gas distribution stations, can now offer their cables to third parties for data transmission services.
The keystone of the radical change of the Monitoring and Remote Control functions lies, in fact, in the change of system philosophy: one no longer talks of management of data coming from separate modems and aimed at "local" plant control but of the fact (decidedly more strategic) that the incoming data are processed in real-time so that they can be integrated into a general business model, a model that will ensure that these data are used in a "multi-service" function. In this spirit Automa can offer specific products to "multi-utilities" companies for structural and environmental monitoring and for security/safety (i.e. access control and fire prevention systems and so on) strictly linked to production and to the specific level of automation (industrial or otherwise).

CAM: machines and processing without secrets
Automa is currently developing specific software for simulating the processes performed by machine tools (once the user has set up the design of the machine and the piece to be produced) even if, from the strategic point of view, in perspective, the possibility of integrating the machine design and the "on-board machine" (namely numerical control) is currently being studied.
In facing the complex sector of machine tools, the company from Pedrengo has also taken on the commitment of "virtual" development of the machine tool, which also requires the performance of a series of operations typical of the CAD sector due to the presence of an editor whose function is to "draw" the solids (i.e. the details constituting the machine elements) that, once assembled, will form and shape to the machine itself. In this way the development of prototypes can be reduced considerably, with an equally significant reduction of costs.
The system perfected by Automa is modular and each module performs determined functions starting from the development of the machine elements (e.g. a trolley, namely a machine element that performs a linear movement and can carry a spindle to which the piece or the tool that carries out the process can be fitted) to end with the assembly of the machine elements and then the finished product, namely the actual machine. The first phase is thus completed, after which the machine is transferred to the area of the "manufacturer" which will personalise it according to the requirements of the end user.
A basic feature of the system is its ease of use: the solids are draw in a natural and facilitated way, after which they can be combined together in numerous ways with a procedure that is totally without complications. Once the machine element has been "created" and after having inserted a suitable set of parameters, there is obviously also the possibility of simulating movement using animation.
It is clear that this "Automa product" is not a design system but, according to Fabio Ghislandi, CAM project manager, "assembly by simulation". The single components can be drawn internally or be imported directly from sized drawings previously produced by the technical office.
The program naturally has other additional movement and animation functions. The next step (still currently work-in-progress) consists of completion of the "tool" part that goes from editing the rough piece to the design of the processes that combine in the production of the finished part.

EPS: optimisation of an idea
Speaking of EPS one finds oneself in the true area of integration projects, namely projects that originate from the scheduling coming from the ERP planning sector: in this regard the Automa commitment consists of the development of systems located downstream of scheduling. Based on the scheduled works orders, an EPS level application distributes the different tasks to the various machines on the same plant. The same application subsequently receives information from the latter relative to the movements of materials and the processes performed that will then feed the company databases to permit the processes to be traced at any time (e.g. for production batches).
In this case the setting up of an integration system goes far beyond the development of a software package, since it does not only involve putting together complex company skills that are very far apart, as well as different problem areas and cultures, but the organisation of a coherent data flow for everyone must be ensured, with precise limits of responsibility for the various data for the various company functions (information systems, programming, production, maintenance and engineering as regards automation).
Meeting the above challenge by offering a product is almost never the solution to the problem: to start with, in fact, it is fundamental to specify the customer's problem in all its details and, obviously, it is equally fundamental to identify its solution while technology comes last, namely it represents the means by which the objective is attained. So the product moves into second place until the time when that product is chosen because it is "the solution" for that given project.
The philosophy regarding integration projects is relatively recent since the awareness of the advantages that can be derived from it has only developed recently.
A business model that plans for full-field integration can be optimised over a period on the basis of real-time checks (this concept must obviously be applied case by case, which means a time period that can go from a day to a millisecond). It is thus possible to promptly meet any request and this is indeed the case, according to Mario Borali, in which one can speak of true integration with the system modelled on the basis of customer requirements.
At this point one cannot speak of integration among levels as an amalgam of different cultures without taking account of the specific time constant that typifies the concept of real-time for each level. For example, at HMI level, an operator can often assess a problem, understand it and solve it in a few minutes (his "real time") while, at ERP level, if a system intervention is to be made a complete team must be assembled consisting of a certain number of people each having different horizontal specialist skills (finance, control, management, logistics and so on): in this case a reaction of a few days can still be considered "real time".
As regards integration at EPS level, Automa has been responsible for numerous applications both operational and in course of completion in Italy and abroad for the most diverse sectors such as, for example, that of coffee and of printing magazines.

SPECIALS: "we don't set ourselves a limit"
The title itself is very clear.
Automa, in fact, aims at attaining a level of experience and skill that will allow it to take any type of application into consideration: this must obviously be understood as a decision of the Pedrengo company tending towards maximum willingness to tackle specific problems of any type.
Among the many cases already undertaken one can mention, for example, establishing the position of the ETR 440 on the rail network, which means being able to establish, at any time, where along the line the train is and at what speed it is travelling; the collection of data about the reaction of given materials subjected to high speed in the centrifuge with subsequent explosion (geotechnical centrifuge); an experimental system for reinforced concrete ecography using an ultrasound device connected to a PC, which consists in assessing the condition of the wall structure without any physical operations such as taking core samples; the "meteor burst" project, a sophisticated idea born before the arrival of satellite transmissions and the Internet, able to exploit the ionisation of the atmospheric layers during the passage of heavenly bodies so that very low power signals can be transmitted, for which Automa developed the data traffic management system.

What future do SCADA systems have?
From the above one deduces that the Automa experience is concentrated along lines where, as Mario Borali says, there are strong and simultaneous synergies so that all these "components" together contribute to a large common base from the point of view of software engineering and technology (the picture of the hand).
This is a new interpretation of "doing automation" only made possible by the fact that the technologies have made this technological convergence possible, so that sectors that are apparently very far apart then find themselves "magically" close together: this is the secret of integration.
To conclude, it is interesting to focus on the Automa point of view as regards SCADA and HMI systems: as of today, in fact, the concept of SCADA and HMI are equivalent, unlike the far too recent past when the two systems represented different entities, given that HMI basically indicated a simple terminal that, compared to SCADA, did not have the capacity of "archiving" the data.
In any case, even if SCADA is currently trying to raise itself up to the EPS level, the products that are regularly put on the market still suffer from the mark of the SCADA producer so that, if the problems relative to the EPS level are not only resolved on the basis of the HMI culture, the suppliers of SCADA systems have difficulty in placing their product since it is "overabundant" for HMI applications and "lacking" for EPS applications. SCADA, although able to communicate generically with a database, cannot operate efficiently with database solutions at Enterprise level and of transactional type.
Indeed, based on these considerations, Automa has decided to distance itself from SCADA systems and decidedly bank on HMI and on integrated systems. In fact, for example, a SCADA system could be adequate if it is required to transfer the data from a process directly to the production manager's office, but when it becomes necessary to integrate these data into a company database where a high level of scalability, solidity, adequate balancing of loads and three-level client-server architecture are required, SCADA, especially the one based on Windows, is no longer sufficient and would thus constitute the weak link in the data supply chain, since the reliability that is now indispensable for the integrated information systems that are currently the basis of modern business models would be lacking.
Site Map Site Map Top of page Top of page