New tools to raise plant performance levels | Calculation and analysis of key performance indicator (KPI) to boost machine and plant productivity is the most widely sought tool off late by many manufacturers. In many cases, manufacturers do not see any further possibilities of improving production processes. Also, due to constant efforts towards improving productivity, most of the manufacturers are engaging into ‘continuous improvement processes’ wherein many machines and plants have already reached their limits of performances or so it seems. But the consistent evaluation of KPI/KPIs often reveals a different story, casting light on additional potential beyond what were believed to be limits of performance.
Rational production is essential for an economically operating business. In this respect, there are a whole range of KPIs that document how efficiently machines and plants are operating. They provide information about where further optimisation measures may yet be advisable. If they are consistently implemented, the productivity of a plant can be raised further.
Normally, the procedure starts by determining the overall equipment effectiveness (OEE). This is defined as the product of the KPIs such as availability, performance and quality with a value range between 0 and 1 or between 0% and 100%. In order to approach the maximum, each of these KPIs must be checked. If any reasons for deviations are determined, measures can be worked out for targeted improvement of the OEE. In this way, gradual improvements can be made to the productivity.
A decisive factor for quick success is ‘automated recording’ of operational data. There are tools available to support such requirements like WinCC/Performance Monitor from Siemens. This tool can be deployed optionally in addition to the SCADA system like WinCC. It enables KPIs to be determined and analysed quickly and easily.
This tool accesses the process data already present in the visualisation system for the calculations to be implemented for various KPIs and OEE. According to the principle of ‘never change a running system’, the required KPIs are configured during operation in the visualisation system. Time-consuming and expensive adaptations at the control level are not necessary. The calculation formulas that are used for the KPIs can also be defined during ongoing operation to facilitate uninterrupted deployment of performance monitoring tool.
All plant components to be monitored (For example, machine modules, plant sections, production lines) are defined as logical units of equipment. The KPIs are assigned to the units of equipment and linked to the existing process data in the SCADA to further start the computation. Once deployed, the comparison of the defined KPIs or similar units of equipment in a plant is possible. If this logical plant model is extended to several factories, findings about the equipment efficiency of entire companies are also possible.
Further, it is possible to assign a plant-specific context to each KPI. This may be, for example, identification of a batch or an assignment to a supplier. Deviating production or quality results allow conclusions to be drawn regarding these production relationships.
This complete transparency of the production processes is absolutely essential for raising productivity. Within the scope of the analysis functionality, such performance monitoring tools can be used on a targeted basis for determining weak spots or recording the details of unwanted process behaviour. In this way, the potential for optimisation, even of complex production processes, can be raised with software support.
In industries, quite frequently, wherein serial machines are used in manufacturing process, capturing standardised machine status information is a very important requirement. For this purpose, a data word with defined states is specified for each machine. Standardised ranges of values simplify not only the engineering but also the subsequent performance analysis.
The advantage for machines of the same type is clear: the time required for commissioning can be significantly reduced, and the subsequent comparison as part of the performance analysis is simplified. A good performance monitor tool also supports standardisation. In contrast to the previously described procedure, the link to the process consists solely of a single structured SCADA process value. So machines with an identical structure can thus be configured for the KPI recording with minimum effort in a very short time.
KPIs are used differently according to the target group within the company, such as plant operators, maintenance personnel, or management. For this reason, it is crucial that all the information is available on a target, group-oriented basis directly on site at the plant and additionally at any location via the corporate intranet or the web.
With such performance monitoring tools, the detailed calculation and evaluation of KPIs is implemented optionally within the scope of the SCADA visualisation. Clear indications show the progression of machine states over any period of time (for example, day, month, shift) and support the comparison of KPI with the analysis of their causes. Machine operators are notified by means of KPIs directly on the process screen. By integrating KPIs into the visualisation system, messages can be generated on violation of a defined limit, or level of quality, performance, or availability.
Apart from the on-site display of this information, flexible access is crucial, in particular for quality assurance and management. This group of people is responsible for production analysis and subsequent optimisation measures. The performance evaluations are then frequently required in offices at some distance from the production halls.
For this reason, the performance functions must be available remotely. The company management is additionally kept up to date with web-based production reports and e-mail messages. Therefore, such performance monitor tools support the processes throughout the company.
The detailed calculation and evaluation of KPIs takes place in three different ways:
- The table control lists chronologically all data and context information used for performance analyses. If necessary, these values can be corrected later.
- The Gantt chart represents the sequence of time-based states (e.g. of a machine startup, tooling, maintenance, operation, fault) of the equipment units as bars along a time axis. This also enables several equipment units to be displayed in parallel, which simplifies the direct comparison of operating states.
- The actual analysis takes place in the performance control. In the bar chart presentation, information can be grouped according to performance units (equipment units, context, or KPI). Individual filters, the option of storing comments, and the flexibility of the presentation all support efficient working. For direct comparison, different time windows (e.g., day, shift) can be arranged side-by-side or overlapped.
The analysis of KPI is completed by drill-down functionality. The initial values of the KPI calculation can be displayed simply by double-clicking on the KPI. This information enables conclusions to be drawn about weak points in the plant. As part of the continuous improvement processes, optimisation measures must be derived from this information, whose success is then evaluated in future performance analyses. This ongoing optimisation process results in a gradual approach toward the maximum overall equipment effectiveness.
Milind Kulkarni
Head, Business Development
Factory Automation, Siemens