Reporting and Analysis Techniques in MES Systems

What kind of data do you record in MES systems?

Every meaningful data in the production area can be recorded. The requirements are determined within the conceptual scope.

Classically, your performance, quality, availability values, which are sub-parameters of your productivity in MES Systems. Your planned or unplanned maintenance times and all maintenance data for your MTTR-MTBF values, which are important KPIs.

While data such as critical hardware lifetimes and diamond tips are recorded, the types of data to be recorded are increasing as the processes evolve because they must always be more dynamic, faster and of higher quality.

Today, dynamic data is also recorded with MES Systems. In other words, while the product is being produced within the infrastructure, what is the temperature in my machine, what is the ambient humidity, did my pressure value in vulcanization stay within the target / planned range while working on a specific work order? Or the tolerance with which the recipe information was followed while the Mix was being prepared is important information and needs to be tracked. Traceability, LOT tracking, cycle time, quality values can be monitored both instantly and retrospectively as the raw material comes to the production site and turns into semi-finished products in each of the sub-operations.

Can we talk about Big Data in MES systems?

What is big data, what should we understand from it? This should be mentioned first. The components that make up big data are;

Velocity: Refers to the speed at which data is generated. Think of social media applications - every like is data, or every signal generated by sensors is data.

Variety: The types of data produced.

Volume: The main problem with big data. As data accumulates, there will be a need to increase capacity.

Variability: It is the transformation of data from one format to another format. It can be exemplified as generating text from a QR code image or format conversion of image files.

Reliability: Verification of stored data.

Value: The data received in its original form has a low value compared to the volume. Only by analyzing the data will it be valuable.

There is no doubt that volume is an important component, but when we talk about Big Data, we should understand the variety of data rather than the amount of data stored.

Structured data: Data whose model can be predefined, whose input format is fixed, and whose storage is easy to query. Database tables and spreadsheets are examples of this. That is, data that we can easily access with row and column references.

Semi-structured data: Similar to structured data, but with a more flexible content. Each new data contains its own model and adds additional hierarchical fields, e.g. XML, JSON.

Unstructured data: Data that does not have a fixed format, such as documents, images, video e-mail messages. We can only make unstructured data meaningful by creating metadata with information such as file name date source.

Data is not only entered into the system through user interfaces, there are various data collection devices included in the system (Sensor, Panel, OPC, PLC devices), or data is also generated through data / reports generated by the warning system.

We can say that data diversity and multiplicity is a natural part of MES.

Is it possible to make Artificial Intelligence - Machine Learning applications with the Big Data obtained? Can you give an example?

Before this topic, we should first talk about what data is and its importance. To put it simply, data (we can also say information) is the most valuable thing that has ever existed since ancient times. When asked what is the oldest known database in the world to date, many people are likely to say Lotus.

In fact, although it may seem so, the first database is the genetic database. DNA and RNA are the best examples. We and other living things live thanks to this information. Data ensures the continuity and survival of businesses. The more real data your data consists of, the more real your predictions will be and will allow you to move forward accurately without deviating from your goals.

We can say that the one who holds the information has the power and data is vital. But collecting data is only the first step.

The next stages are the stages that will make sense of your data and are very important. When we say a developing system that continuously learns by processing the collected data through certain models, we enter the world of artificial intelligence-machine learning applications. This is such a wide-ranging subject that it can be used in every sector, everywhere and develops the area where it is used at an incredible acceleration. Instead of decisions that will take days of analysis, you can create a system that can make decisions in seconds and suggest the right path.

In which areas can it be applied on MES?

First of all, automatic bottleneck detection should be mentioned. A system that finds the bottleneck points in production will alert you in the early stages of the bottleneck and enable you to take quick action on this issue. In this way, instead of learning through reports received after production, you can have information without losing time.

Secondly, it is used on speed suggestions. The system will give you the most suitable speed by looking at your previous work history with a certain algorithm. This can be thought of not only as the most repeated speed of a product's operation at a station, but also as adding parameters such as the speed with the least waste, the speed with the least malfunctions.

Machine learning is being developed in many areas. For example, when we add the probability of failure of an equipment used in production with the production made on a raw material or the speed parameter, it paves the way for a structure that can prevent many failures.

As human beings, we are on the verge of a great technological development with Industry 4.0 in the future artificial intelligence and machine learning. These are just the beginning. Incredible technologies to develop and discover await us.

What kind of reporting is possible in MES systems?

Shift, daily, weekly, monthly and annual reports,
Reports by operator, machine, line, plan, country,
Customized reports based on the data collected,

Reporting of the personnel who experience the most downtime and their reasons

What kind of analyzes can we talk about in MES systems?

The most used analysis in MES systems can be called OEE analysis. It allows us to obtain information about the instant or periodic course of the factory, work center, line, station. It is the first stop on the way to the root cause. According to the values that come as a result of your analysis, you determine your improvement area and move to the next step on the way to the root cause. For example, if you want to examine in detail a loss that affects usability related to a pause, your 2nd step will be posture analysis. However, if you have a loss that affects performance, your 2nd step would be cycle time analysis or cycle time pareto histogram. If your loss is due to quality, quality analysis will guide you as your 2nd stage. In addition to all these, trend analysis will guide you when you want to follow and analyze trends in the relevant KPI value (Availability, Quality, Performance, OEE).

The benefit provided by the improvement after the action taken regarding the loss you made trend analysis can be clearly calculated in advance and can be followed after the action. These are the main analyzes used in MES Systems and it is possible to integrate different KPI calculation models used or developed by companies into the system and create new analyzes. MES Systems constitute the backbone of corporate data management in the enterprise. As this structure develops, it both facilitates our work with its incredibly beneficial features and allows us to recognize ourselves as a business.

Can you tell us about automated reporting services and their importance?

If the goal is to maintain stable production, instant alerts or scheduled reports will be needed. Since data analysis directly affects the decision process, reports and notifications need to be produced at the right time with the right content. The right time and the right content of course depend on system resources and report design. Therefore, it may be necessary to study the field well for configuration at optimum values and design new reports instead of ready-made template reports. Optimizing reports should be easy to redesign.

Of course, it is also important how the reports produced will be distributed and how they will be presented. Because they are easy to read;

Generally in PDF format,
HTML format,
JPEG image content,
The warning system can also be requested in the form of a simple warning text.

Reports

Alerts generated by the system, which can be transmitted via e-mail,
Via SMS,
It should also be able to be pushed to mobile devices.

It will enable people in the decision-making process to take action faster. Therefore, it can be said that the reporting service has a very critical importance.

As in every system, flexible reporting and the ability to design reports on your own are extremely important in MES Systems, can you tell us a little about this subject?

If data with a wide variety of content and files in different formats can be summarized in an organized way, it will produce meaningful new data. Isn't this what we call a report? In systems with Big Data, it is not enough to produce reports with SQL queries only.

In addition to the database server, data generated by other sources should be able to be reported, and unstructured data, such as an image or any document, should be easily reported.

Instead of static reports, it is expected to be conditional, reprogrammable when necessary, that is, dynamic and support certain scripting languages.

There will be a need for tools with user-specific customizations, understandable, user-friendly interfaces, where even those who know some basic concepts can easily get involved. Not only reports but also Andon Dashboard designs should be possible.

To summarize, -flexible reporting-, it is important to be able to report the available data without difficulty and present it in up-to-date and common formats.

Is it possible to make integrated reports with lower and upper systems in MES systems?

Case studies like this have been done in many enterprises. When technologies are integrated, the algorithm of integration is the only limit for you. It is possible to produce rich integrated reports on every area you are integrated with. In this way, you can compare the planned theoretical data with the reality and shed light on how the next planning should be done.

Another issue is that it allows your cost analysis to be done on reality. Cost analysis is usually calculated as lossless, but in the field the situation is not so constant. Losses in production affect the cost of an order very seriously. However, by looking at the statistics of past productions, it allows the business to make more real cost analysis, to give more real tact time to its customers, to give more accurate prices by calculating the gross production time cost.