MES (manufacturing execution systems) have emerged out of a mish-mash of applications that were developed for different organisations to aid them with specific aspects of their business. As time went by, the capabilities of these programs became more understood and appreciated - and their use became more popular. Data formats, etc gravitated toward more universally appropriate standards. Microsoft's monopoly on software for PC platforms naturally meant that Microsoft data formats and protocols have been adopted. Successful software houses began acquiring some of the most useful modules and bundling them into suites that are the MES we have today. Many enterprises are now hankering after the ultimate enterprise's dream.
Getting the pie out of the sky
Perceived by many to be pie in the sky stuff, MES often takes on the appearance of a bunch of buzzwords orbiting lofty concepts that never leave the back of the cigarette box. The dream is ultimately a central nervous system, so that when any part of an organisation itches, management scratches. Perhaps full-blown MES could be seen as the main part of a central nervous system - the fully-functioning spinal column of an organisation. It sits between the ERP and scada layers and is able to assist management in ensuring that every part of the enterprise is running lean and strong. You hear people talking about it, but would be hard-pressed to find any commonplace evidence of practical (and complete) MES that the average businessman would happily adopt. The perception is that it has just not taken off in the big way that was expected.
MES is with us
Make no mistake, manufacturing execution systems are certainly real and with us - and they are used to great effect by organisations that properly implement them. The impact they have on a business's productivity can be dramatic. MES implementations (even small-scale implementations) have been known to pay for themselves in comparatively little time. Early South African adopters of 'full-blown' MES include the likes of Sasol and the then-called Alusaf.
So what is the problem?
MES is essentially all about housekeeping the logistics of an array of concepts that differs, not only from one organisation to another, but also changes over time. Different enterprises will naturally partition their functionality differently, and changes in the market, mergers and splits will have significant impact on how best the array of concepts needs to be configured for the organisation and its circumstances.
In reality, many businesses are not so complex as to require a full-blown, intelligent software model of their enterprise. For the simplest businesses, things are so glaringly obvious that management has no need of any software to point out these things to them. As complexity increases, it becomes harder and harder for a human to be able to take all into consideration - and know how simple changes in certain areas could make a difference that may determine whether the company retrenches or expands.
Different things to different people
Different software houses have different beliefs regarding just what exactly constitutes MES. After chatting to some of the biggest names in scada and MES in SA, it becomes apparent that MES is really an attempt at acquiring software to mimic an employee, or team of employees, possessing the super-human abilities of never forgetting, never overlooking and always unashamedly presenting the necessary naked facts in a form that permits prompt and advantageous decisions to be made by a person with the authority to take the necessary action to optimise the business's effectiveness. Implementing MES for a complex enterprise is vast task. So 'full-blown' MES begins to resemble that proverbial can of worms. It appears to defy any universally agreeable game plan. Many throw their hands up and say that it is an idea ahead of its time.
Building MES in simple terms
Let us look at an analogy that hopefully should clearly illustrate the dilemma that we face. Imagine, if you will, that our enterprise is a big, fancy house. The bricks are laid by means of programming languages (like C, etc). Applications that form part of MES suites are fully fitted, slightly flexible, ready-built rooms.
Obviously the house may be built exactly the way we would like it if we used bricks. But this would be expensive to modify, and with time scarce and builders demanding high fees, it becomes far more cost-effective to make use of the selection of ready-made rooms. This is well and good if the rooms contain the fittings that we need, have the doors in the right place and are suitably dimensioned. The problem is that many of these rooms are derived from designs that were custom jobs (made with bricks) for others who were not necessarily intending to assemble a house like ours.
Where businesses are similar, the houses are similar and they may be built from a similar set of rooms - this is seen to be true in pharmaceutical enterprises, where MES for different pharmaceutical companies has a great deal in common. This commonality does also exist between companies in other similar fields.
As time goes by the industry is slowly accumulating a library of room designs and establishing which are best suited to which final house requirement. It is more complicated than this, as the interconnecting of the rooms and setting the house in order is still a mammoth task.
So the dilemma is that bricks are too basic and builders cost too much. Pre-fitted rooms are too specific, and often do not even faintly resemble what we think we need. It would appear that the answer lies somewhere between. We want a system that makes use of house components that are simple enough to understand almost instantly, flexible enough to be used to make any conceivable room, but still powerful enough to perform significant tasks - and suitable for DIY.
A different spreadsheet package for every industry?
People from many different disciplines make very effective use of spreadsheets and word processors. There is no need for these two versatile applications to be produced in different versions for all the different disciplines. It makes sense that once MES has come of age, it should be sufficiently flexible, expandable and understandable to be installed, set up, and maintained by its users. Of course the initial linking to the ERP and scada levels would likely require the services of an able Systems Integrator, but after that, any changes within the MES should be easily performed by the user. Ideally there would be a standard scada interface, a standard ERP interface, compatibility with word processor and spreadsheets, and a user interface that can work on any browser.
There is a need
Many involved in the MES development and vending business feel that there is still a big lack of connection between ERP and the control layer and they are pushing for industry to make use of MES. In the more complex enterprises this is certainly true. As MES Suites become more friendly and versatile, smaller and smaller enterprises will begin adopting them. At the start of the industrial revolution, efficiency was almost a non-issue, but now, with world-wide competition, tighter restrictions on environmental concerns, tighter restrictions on the personnel issues ... this all eats away at any 'slack' that would not have been an issue in days gone by. The line between flourishing and floundering is becoming quite thin.
Full-blown MES is something to empower and be used by top management ... it is not something that is sold to the engineering layer people. Typically for it to work there needs to be 'buy-in' from top management right through the engineering layer.
Upper management are no strangers to software packages. They are familiar with accounting software and office suites. But these packages tend to be clearly defined and follow popular and familiar guidelines. Perhaps the closest thing to MES that these folk get to play with is ERP. The good ERP suites are designed around recognised best practice business principles, and as such are fairly predictable in their structure and methodology. The user can click through the options and 'fill in the blanks', thus setting it up to run in a way that is useful to their business. This is an oversimplification, but compared to MES, this is fair comment.
With MES suites, whew ... Here the user is faced with a whole bunch of 'sub-applications' that were of great use to someone else, and are not familiar at all. It is certainly not a matter of filling in a few company particulars and clicking on some options. Currently full-blown MES is assembled and configured on a case-by-case basis. It can take a year or more just to define the functionality that is needed. Viewed in one hit it is overwhelming and unfamiliar.
Different software houses and vendors can be passionate about their differing views on what MES is. My personal opinion is that if a bucket is full to the top with water, or if it has a little in the bottom, the stuff sloshing in the bucket is still water. So - depending on the needs of an enterprise, they may need a very thorough MES set-up, they may need no MES at all - or they may need anything in between. At the moment it appears that there is something available for everybody.
For those only requiring control layer reporting, downtime analysis and the like, scada providers provide excellent tools that are becoming very easy to use. For those wanting to install all the bells and whistles, thoroughly interconnecting all parts of the organisation, there are companies that are ready to help too.
John Gibbs, Editor SA Instrumentation & Control, 031 764 0593, [email protected], www.technews.co.za
Search for MES, scada and system integrators at: www.ibg.co.za
© Technews Publishing (Pty) Ltd | All Rights Reserved
printer friendly version