System to and method of monitoring condition of process tool

• US 20050256601A1
• Filed: 02/07/2005
• Published: 11/17/2005
• Est. Priority Date: 05/13/2004
• Status: Active Grant

• Current Assignee: Samsung Electronics Co. Ltd.
• Original Assignee: Samsung Electronics Co. Ltd.
• Inventors: Lee, Seung Jun, Doh, Seung Yong, Park, Chung Hun, Jang, Yoo Seok
• Granted Patent: US 7,194,325 B2
• Time in Patent Office: Days
• US Class Current: 700/108
• CPC Class Codes: G05B 19/4184 characterised by fault tolerance, reliability of production system
• CPC Class Codes: G05B 2219/31357 Observer based fault detection, use model
• CPC Class Codes: G05B 2219/45031 Manufacturing semiconductor wafers
• CPC Class Codes: Y02P 90/02 Total factory control, e.g. smart factories, flexible manufacturing systems [FMS] or integrated manufacturing systems [IMS]

Abstract

A system to and method of monitoring a condition of a process tool. The system monitors a condition of a process tool to correctly detect a faulty operation or malfunction of the process tool. The system to monitor the condition of the process tool includes a first model storage unit to store one or more good models generated by data associated with the process tool, a second model storage unit to store one or more faulty models generated by the data associated with the process tool, a model selector to receive tool data from the process tool, and to select one of the good models and one of the faulty models in association with the received tool data, and an error detector to receive process data from the process tool, to compare the received process data with the good and faulty models selected by the model selector, and to estimate a condition of the process tool.

Claims

1. A system to monitor a condition of a process tool, comprising:

a first model storage unit to store one or more good models generated by data associated with the process tool;

a second model storage unit to store one or more faulty models generated by the data associated with the process tool;

a model selector to receive tool data from the process tool, and to select one of the good models and one of the faulty models in association with the received tool data; and

an error detector to receive process data from the process tool, to compare the received process data with the good and faulty models selected by the model selector, and to estimate a condition of the process tool.

2. The system according to claim 1, wherein the one or more good models are generated by various good data indicative of a normal condition of the process tool, and the one or more faulty models are generated by various faulty data indicative of an abnormal condition of the process tool.

3. The system according to claim 2, wherein the good data and the faulty data are differently determined according to a variety of classification references.

4. The system according to claim 2, wherein the one or more good models are modeled by different modeling methods when there are two or more good models.

5. The system according to claim 2, wherein the one or more faulty models are modeled by different modeling methods when there are two or more faulty models.

6. The system according to claim 2, wherein the error detector determines whether the process data is contained in the good model and the faulty model, and determines whether the process data is similar to the good model and the faulty model, such that it determines a condition of the process tool according to the determined results.

7. The system according to claim 6, wherein the error detector determines the condition of the process tool to be a normal condition when the process data is contained in the good model or is more similar to the good model than the faulty model, or determines the condition of the process tool to be an abnormal condition when the process data is contained in the faulty model or is more similar to the faulty model than the good model.

8. The system according to claim 1, wherein the tool data includes a variety of information including a degree of obsolescence of the process tool, presence or absence of replacement of a component in the process tool, and presence or absence of a process change.

9. The system according to claim 1, further comprising:

a reference database (DB) to store the good data and the faulty data.

10. The system according to claim 1, further comprising:

a display to display the result of monitoring the condition of the process tool.

11. A system to monitor a condition of a process tool, comprising:

a first model storage unit to store at least one good model indicative of reference good data for determining a condition of the process tool;

a second model storage unit to store at least one faulty model indicative of reference faulty data for determining a condition of the process tool;

a model selector to receive tool data from the process tool, and to select the good model and the faulty model in association with the received tool data; and

an error detector to receive process data from the process tool, to compare the received process data with the good and faulty models selected by the model selector, and to estimate a condition of the process tool.

12. A method of monitoring a condition of a process tool, comprising the operations of:

storing one or more good models generated by good data collected in a normal condition of a process tool and one or more good models generated by faulty data collected in an abnormal condition of the process tool;

selecting a good model and a faulty model in association with process tool data received from the process tool;

comparing the process data received from the process tool with the selected good model and the selected faulty model; and

determining a condition of the process tool according to the result of the comparison.

13. The method according to claim 12, wherein the condition of the process tool is determined to be a normal or abnormal condition according to a result indicating whether the process data is contained in the good model and the faulty model and a similarity result indicating whether the process data is similar to the good model and the faulty model.

14. The method according to claim 13, wherein the similarity between the process data and the good model and the other similarity between the process data and the faulty model are weighted with different weights if the condition of the process tool is determined according to the similarity result, such that the similarity between the process data and the good model and the other similarity between the process data and the faulty model are calculated.

15. A method of monitoring conditions of plural process tools, comprising the operations of:

storing one or more good models generated by good process data collected in a normal condition of each process tool and one or more good models generated by faulty process data collected in an abnormal condition of each process tool;

selecting a good model and a faulty model in association with tool data received from each process tool; and

determining a condition of the process tool based on a comparison between process data received from each process tool and both the selected good model and the selected faulty model.

16. The method of claim 15, further comprising determining whether the process data of each process tool is contained in the respective faulty model if the comparison does not determine that the process data of each process tool is in the respective good model.

17. The method of claim 16, further comprising:

calculating a similarity between the process data of each process tool and each of the respective good and faulty models if the process data of each process tool in determined not to be contained in the respective faulty model; and

determining whether each process tool is in the normal condition or the abnormal condition by determining whether the similarity between the respective process data and the respective good model is greater than the similarity between the respective process data and the respective faulty model,

18. A system to monitor a condition of plural process tools, comprising:

a model storage unit to store one or more good models generated by data associated with each process tool and one or more faulty models generated by the data associated with each process tool;

a model selector to receive tool data from each process tool, and to select one of the good models and one of the faulty models in association with the received tool data for each process tool; and

an error detector to receive process data from each process tool, to compare the received process data of each process tool with the respective good and faulty models selected by the model selector, and to estimate a condition of each process tool in accordance with the comparison.

19. The system according to claim 18, wherein:

the one or more good models are generated by various good data indicative of a normal condition of the process tool, and the one or more faulty models are generated by various faulty data indicative of an abnormal condition of the process tool; and

the good data and the faulty data are differently determined according to a variety of classification references.

20. The system according to claim 18, wherein the error detector determines whether the process data is contained in the good model and the faulty model, and determines whether the process data is similar to the good model and the faulty model, such that the error detector determines a condition of the process tool according to the determined results.