Modern manufacturing is increasing the use of machines, robots, various sensors, and other devices, which systematically collect data and possibly store them in a way that can be used later for analysis. If we add a modern MES system, which stores data about workers, standstill, operation of machinery and manufactured pieces, we have valuable data, which is often unused. One of the main motives for introducing MES system is availability of real time and history data about wide range of production measures.

Challenge

When analyzing operations data, we had to answer the following questions with  TIBCO Spotfire® solutions:

• General questions on manufactured quantities, standstill and similar.
• What is the performance of individual workers?
• How working hours are used?
• Does the number of individual operations carried out by a worker impact his/her productivity? How is performance dependent on shifts (morning, afternoon, night)?
• What is the performance of machinery / workers during a shift?
• OEE and its components values (availability, performance, quality)?
• What is the relationship between the preparation and operation of machines?
• Is the sequence of the working orders optimal?
• What is the degree of completion of each work order?
• What are the reaction times of maintenance?
• Where was the largest cost of poor quality generated?
• Are operations standard times appropriate?

With operations analytics we do not want to study only the basic measures, such as manufactured quantities of waste but we want to carry out so-called root cause analysis, i.e. the analysis of the root causes of deviation from the expected values.

Solutions

TIBCO Spotfire® solutions were in the production of analysts because of its features, such as:

• Direct link to information resources, thereby studying the last possible value and taking immediate action.
• Visualization of large amounts of data over a larger period of time high number of workers, machines, products and work orders.
• Large volume of data that can not be studied with the usual tools.
• Ad hoc analytics and user-friendly interface for the end user – it is impossible to predict all the questions and answer them with pretty dashboards. This is precisely the essence of the so-called “Root cause” analysis. If this were not the case, the management of production processes would be significantly easier.
• Using floor plans to analyse data. In the example above we used stand still by category and over time. This can be later used for 6 Sigma and similar methodologies. If the methodologies require the use of statistical methods there is a large number of built-in statistical functions and visualization.

How TIBCO Spotfire® was used

Preučevanje zastojev ali drugih proizvodnih parametrov je možno preko tlorisa (naslovna slika), ki tako hitro in vizualno sporoča, kje in kakšni zastoji nastajajo, ter kakšna je njihova struktura. Zastoje je smiselno preučevati po njihovem številu in pa času trajanja. Ob dobri kategorizaciji vzrokov lahko hitro odkrijemo vzroke za nizko produktivnost, nizek OEE ali druge proizvodne kazalnike.

OEE is the most general indicator of the performance in manufacturing which is not particularly useful if you do not know its submeasures (availability, performance and quality) and changes in these indicators over time and across departments.

Here we show an example of performance by departments, workers and in time. With the marking of data (column) we easily drill down to lower levels such as employees, machines or each work order.

By capturing data directly from the machines we get detailed information on how these are  operated during the day. Red dots indicate a standstill, gray, however, they indicate normal operation. gives us quick insight into performance potential with or without disturbances. Often we discovertha performance is relatively high during the shift with trend of decreasing towards the end of working day, although there is no specific cause.