Why does a Stocker (STK) system require condition monitoring?

Stocker (STK) systems are responsible for storing and transporting work-in-process products, finished goods, and cassettes in semiconductor and display manufacturing facilities. As a key component of the AMHS infrastructure, unstable handling operations, motor abnormalities, or motion deviations may reduce transport efficiency, compromise product safety, and affect production uptime.

What operations are typically monitored in Crane STK cassette handling?

Crane STK cassette handling typically includes Take, Put, Go, and Come operations. Take refers to cassette pickup, Put refers to cassette placement, Go represents movement toward the target storage location, and Come indicates the return movement. Monitoring these patterns helps ensure stable handling performance.

How does VMS-ML monitor Stocker systems?

The VMS-ML Machine Learning Monitoring System learns normal Crane STK operating behaviors and establishes standard action models. Real-time signals are continuously compared against learned patterns, allowing the system to identify abnormal actions or unstable operating conditions.

What does a dynamic similarity score represent in STK monitoring?

A dynamic similarity score measures how closely real-time signals match learned standard operation patterns. Under normal conditions, pickup and placement operations can achieve average similarity scores of approximately 91% to 92%. Lower similarity scores may indicate unstable movements or abnormal behavior under load.

What can motor spike anomalies indicate in an STK system?

Motor spike anomalies may indicate unstable motor performance, abnormal mechanical loading, inconsistent axis movement, or equipment degradation. In this case, the L2 motor exhibited spike anomalies and significant deviations from healthy motor patterns, resulting in warning alerts.

What are the benefits of implementing STK monitoring?

STK monitoring enables a single sensor to track multiple axis movements, detect motor spikes, identify inconsistent handling operations, and reveal stability degradation under load. Combined with VMS-PH Dynamic Analyzer diagnostics, it supports predictive maintenance and helps prevent unexpected downtime.

Case

Application

FAQ

Support

How to Measure Stocker (STK) Storage Equipment?

Case | How to Measure Stocker (STK) Storage Equipment?

In the era of Industry 4.0, many automated devices replace manual labor, accelerating processes and simplifying heavy workloads and complex procedures. Stocker storage equipment has taken over critical material transportation tasks. How can measurements be performed to ensure stability during the transportation process?

Stocker Features

In the era of Industry 4.0, automation has increasingly replaced manual labor, enhancing efficiency and simplifying complex workflows. With the advancements in semiconductor and digital technology, silicon wafers are becoming larger for economic benefits, making transportation more challenging.

To address this, automated transportation systems have been implemented within factories, such as the AMHS (Automatic Material Handling System), OHT overhead transportation system, and Stocker storage equipment. These systems reduce operator workload and prevent human errors or accidents. Stockers come in different types and operational models depending on factory needs, including ZIP Stocker, Foup Stocker, Tower Stocker, and Linear Stocker. Most of these designs follow a modular architecture to maximize space efficiency and optimize usage.

Solutions and Monitoring Overview

VMS-ML Machine Learning Intelligent Monitoring System
Stockers are responsible for storing glass substrates in panel factories and storing semi-finished or finished semiconductor products in semiconductor plants.

Gude Technology focuses on monitoring and diagnosing the movement behaviors of Crane STK when handling cassettes and assessing the quality of brake servo motors.

By utilizing the OLVMS-ML machine learning intelligent monitoring system to learn correct movement behaviors as a benchmark, each action is monitored and diagnosed.

The system identifies abnormal or unstable actions in the equipment, enabling predictive maintenance and preventing unexpected failures.

Measurement Status

Crane STK Cassette Handling Process
Learning the sequential movements of the Crane STK Cassette handling process:

Take: Crane picks up the cassette
Put: Crane places the cassette
Go: Crane moves to the innermost position
Come: Crane returns to the starting position
Establishing a management pattern for Take, Put, Go, and Come actions.

Crane STK Cassette Handling Action Recognition

Yellow line: Machine learning movement standard, White line: Real-time signal
The system automatically tracks arm movements, managing both pick and place actions without misidentification.
Normal machine operation average dynamic similarity: Put - 91%, Take - 92%
When the arm operates normally, the machine learning motion (yellow line) aligns with the real-time signal (white line).

Crane STK Empty Run Movement (Without Cassette)

Crane Returning to Starting Position (Come)

Yellow line: Machine learning movement standard, White line: Real-time signal
High stability in dynamic similarity scores during empty runs!

Crane STK Loaded vs. Unloaded Movement Comparison

Yellow line: Machine learning movement standard, White line: Real-time signal
Inconsistent motion after loading causes a drop in dynamic similarity scores!
When the arm operates under load, the machine learning motion and real-time signal become misaligned.

Crane Empty Run Motor Timing Vibration Comparison

Motor Operation Visualization

Crane STK Load Motor L1 and L2 Machine Learning Comparison

Crane STK Load Motor L1

Crane STK Load Motor L2

Measurement Conclusion

Using the in-factory OLVMS-ML monitoring system, it was found that STK L2 exhibits sudden surges, with a significant deviation in similarity scores when compared to a healthy motor pattern.

L1 is affected by L2, showing fluctuations and abnormal frequency. R2 is slightly impacted by L2, showing minor fluctuations compared to R1. Only R1, which is furthest from L2, remains relatively stable.

The VMS-ML Machine Learning Intelligent Monitoring System can analyze the movement conditions of each axis with a single sensor, then use the VMS-PH Dynamic Analyzer to identify the root cause of anomalies.

VMS-ML Machine Learning Monitoring System

VMS-ML

Monitors and diagnoses individual movements.

FAQ

Why do Stocker (STK) storage devices need to be measured?
Stocker (STK) storage devices are responsible for storing and transporting products, finished goods, or Cassettes in semiconductor and panel factories, making them essential equipment in automated material handling systems. If instability, motor abnormalities, or movement deviations occur during the handling process, it may affect transport efficiency, product safety, and production line utilization. Therefore, it is necessary to grasp the equipment's operating status through measurement.

What actions are usually monitored when a Crane STK handles Cassettes?
When a Crane STK handles Cassettes, continuous actions such as Take, Put, Go, and Come can be monitored. 'Take' is the action of the Crane picking up a Cassette, 'Put' is placing it, 'Go' is the Crane sliding to the innermost position, and 'Come' is the Crane returning to its starting position. By establishing management Patterns for each action, it can be determined whether the handling process is stable.

How does VMS-ML monitor Stocker storage devices?
The VMS-ML machine learning intelligent monitoring system can learn the correct movement behavior of the Crane STK as a standard, and monitor and diagnose individual actions. The system compares the machine learning action standards with real-time signals, using dynamic similarity scores to determine in which action the equipment experiences an abnormal or unstable state.

What does the dynamic similarity score represent in STK monitoring?
The dynamic similarity score represents how close the real-time signal is to the normal action standard. When the Crane STK operates normally, the machine learning action standard and the real-time signal align closely, and the average dynamic similarity for picking and placing can reach about 91% to 92%. If the operation is inconsistent after loading, the similarity score will drop, indicating that action stability may have deteriorated.

What problems might an abnormal STK motor surge indicate?
If an abnormal surge occurs in the STK motor during load or idle operations, it may indicate unstable motor operation, abnormal mechanical load, inconsistent axial operation, or a decline in equipment health. In this case, the L2 motor showed an abnormal surge, and its similarity differed significantly from a healthy motor Pattern, prompting the system to generate a warning.

What are the benefits of implementing STK storage device monitoring?
After implementing STK storage device monitoring, a single sensor can be used to understand the operating conditions of each axis, enabling the early detection of issues such as motor surges, inconsistent handling actions, and decreased stability after loading. Subsequently, it can be paired with the VMS-PH dynamic analyzer to troubleshoot the causes of abnormalities, assisting in predictive maintenance and avoiding unexpected downtime.