Ranking right up there with tax audits and waiting for medical results is the stress associated with drawing up the functional specifications for a manufacturing execution system (MES). Most of this stress isn`t due to the hard and meticulous work involved, but rather to the consequences of not making provision for future needs and unforeseen circumstances. As we live in a constantly and rapidly changing world, there`s a good chance that specifications and the system they describe will be out of date by commissioning time. So, what`s the answer?
In most companies, the best way to get a lot of attention is to spend a few million on a system that doesn`t deliver the goods or live up to the expectations of top management. Stating that the needed functionality didn`t exist at the time the original specification was drawn up is no longer an excuse.
While open systems readily lend themselves to modifications and interoperability, changes in code can be costly and propagate errors that can have even more costly consequences. Code changes are usually out of the end-user domain requiring the intervention of experts who know a lot about coding but not necessarily as much about why the code is needed.
New, different or added functionality is required every day by many organisations looking to optimise their efficiency, profitability and return on investment. For example, the area of maintenance management is being paid a lot of attention these days, of drastically increased capital and plant expenditure in a business landscape where customer dissatisfaction is simply another notch on the competitor`s belt.
In most companies, the best way to get a lot of attention is to spend a few million on a system that doesn`t deliver the goods or live up to the expectations of top management.
Michael le Plastrier, director of EOH Consulting Services.
A lot has been said with regard to maintenance, and current maintenance systems have a lot to offer. Those considering a maintenance system at present have a good choice of options either at the plant or enterprise level:
* Preventive maintenance (proactive)
* Predictive maintenance
* Condition-based maintenance
* Total productive maintenance
* Reliability-centred maintenance
While reading through the list, did any one approach seem particularly applicable to your environment?
At one end of the scale, we have a calendar-based approach to maintenance - look at it on the 23rd of every month irrespective of whether it`s melted or not. With reliable plant, this can lead to bored operators ticking the "OK" box for the unit months ahead of its demise.
At the other end of the scale is a real-time environment capturing real information on plant items and predicting when they should be serviced in order to minimise downtime. Who`s to say that the approach you choose will be suitable for and able to keep up with your diverse, changing and probably unique maintenance needs?
Perhaps an evolutionary approach is the answer. "Grow as you go" makes a lot of sense because it caters for changing needs without the trauma of specifying the functionality of obsolescence-prone monolithic systems upfront. But this means that you have to be empowered to be master of your own fate insofar as system changes are concerned without having to become a software house. So, how does one add functionality without adding code?
Rather than diving headlong into complex PLC and application software, imagine that you could simply add these three parameters as attributes to a standard pump template, which you could then deploy at every occurrence of these pumps in your SCADA system. Suddenly, you have "intelligent" pumps that give you the maintenance information you need for analysis by any number of available analytical tools. In one stroke, you`ve specified new data items that never existed before, made these operational throughout the plant and analysed the results with standard analytical software - all without writing a single line of code or ingesting abnormal quantities of antacid.
Intelligence and flexibility don`t have to reside in hard-wired code but in the flexibility of object-oriented attributes. These can be static or dynamic, deriving their values from the attributes of other entities or objects - rather like DNA and genes.
It`s not all pie-in-the-sky stuff. Today, technology is readily available and is already in use by many South African manufacturing companies such as SABMiller, Coca-Cola and SCAW Metals. They are using these technologies not only to control their existing production processes in a unified way, but to allow them to evolve at a pace of their choosing into MES, maintenance and other applications without committing themselves upfront to functionality they don`t know they will need.
In short, the technology is allowing them to add functionality after the fact and to drive their production processes the way they`ve always wanted - in the evolutionary and logical way that best meets their changing requirements.
Share