Monitor machine anomalities in real time;
grasp the opportunity of yield rate improvement.
Best Solution to Massive Data of Machines
The amount of data produced by sensors on machines is massive with hundreds of parameters regarding temperature, pressure, gas, etc. being monitored every second or even shorter. Specially designed system architecture (Hadoop + Spark) is definitely required in order to store and process such massive data.
Sensor data produced during machine operation is used to detect the health status of the process. Opportunities can be grasped at the time of production for rapid and accurate problem identification.
When anomalies are detected, the system will immediately notify the owners responsible for troubleshooting.
Owners responsible can also utilize control charts and relevant analytical data to speed up the identification of root cause and carry out necessary treatment.
When anomalies are detected, the system notifies owners responsible automatically for early treatment. The homepage displays personal abnormal event monitoring lists and process control charts. All the latest abnormal events are listed and can be browsed in a quick and holistic manner.
The system automatically organizes an anomaly list and displays the abnormal parameter of that production unit. All the latest abnormal events are listed and can be browsed in a quick and holistic manner.
Trend charts are plotted in accordance with the production parameter values produced by each machine every second of the day.
Owners responsible can also set up a reference line for warning for continuous production trend monitoring.
Sensor data is produced by production machines 24/7 but not all of them has value. For instance, the resting time or waiting time of a machine is not relevant to manufacturing; removing these invalid data is beneficial for the actual representation of production data.
Manufacturing, although a continuous process, is rather complex in its operation and can be divided into multiple stages. Dividing the process into several segments makes delicate differences more identifiable.
Suitable control indicators can be defined based on the different features along the process. The SPC control chart can monitor the process into several stages more accurately and in detail.
Anomalies are shown on the control charts with a different color so that users can quickly browse through and compare hundreds of control charts on a daily basis.
Normally, manual definitions of process segments can only be done when sensor data changes significantly or trend charts are closely observed. It is, in general, very difficult to define manually a set of process segments that integrate all parameters.
EEAS System utilizes rigorous statistical methodologies to assist users in simulating segmentation conditions and apply them to the setting, reducing the difficulty and uncertainty of manual definitions, and improving the appropriateness of process segmentation.
When the system employed to simulate segmentation conditions, the intended condition will first be applied to profile trend charts to check if the simulation results are appropriate.
The control limit of monitoring indicators can be set from the system. Statistical methodologies are additionally provided for control limit simulation calculation. Users can decide the reference setting on their own.
When anomalies are detected, owners responsible will immediately be notified and be provided with relevant information, allowing them to treat the anomaly is a fast manner and avoid losses. Information such as profile trends and production traceability can be tracked through anomaly-related data, allowing rapid and in-depth analysis of root causes.
Information of EEAS System can be connected with that of EDA/YMS Data Platform. When machine anomalies are detected by EEAS System, information such as the abnormal batch can be easily and quickly linked to the 4M1E engineering data of EDA/YMS System to check immediately whether such anomaly affect product quality and vice versa, issues of the products can also be traced to parameter anomalies of the machine producing such products. This feature greatly reduces the inconvenience caused by data transfer and system or platform switch, streamlining the analytical process for maximum benefits.
EEAS is a data platform of both data integration and data analysis. With just a URL and an authorization, users can access production data and enjoy system functions remotely from all across the facilities.
With the thin client and database solution, users are no longer limited to local hardware and software. Accessing massive amounts of data is just as easy as browsing websites.
The flexible system architecture allows easy expansion of EEAS data platform even when new machine data types need to be integrated to the system when the systems goes online.
The system is easy to deploy, easy to maintain, and easy to expand; and with integrated functions such as access management, system management, and user habit analysis, EEAS meets the needs of IT managers in all aspects.
For data from highly automated machines, data can be collected in real time and be organized to the EEAS data platform in an automatic and smooth manner.
EEAS System requires regular mass data access and computing; therefore, TYNE EEAS has designed the system in a special manner for it to achieve maximum efficiency. The system has high operational speed, high stability, and excellent performance.
