Toshiharu Miwa

Automated stepper load balance allocation system

In order to improve productivity in the photolithography process of high-product mix/low-volume factories, an automated stepper load balance allocation system was developed. The system enables us to maintain a balanced load distribution of tool constraint layers among all steppers. We developed a processing time estimation function and a load balance allocation function. The processing time estimation function calculates accurate loads based on the processing times related to the product, the process layer, and the stepper. The load balance allocation function was realized by applying a dynamic programming method. The system has decreased the deviation in the total processing time among all steppers by 10%, compared to the conventional manual allocation method.

Focus and CD control by scatterometry measurements for 65/45nm node devices

A method using scatterometry for simultaneous focus and critical dimension (CD) control method has been developed. Our focus and CD measurement method uses a five-layer scatterometry model and provides stable focus measurement when the exposure dose fluctuates. We utilize this feature and consider applying it to the response surface methodology model for focus and CD control. This control optimizes focus and calculates the correct dose allowing for the focus effect. We have confirmed that this method controls photoresist shape accurately and reduces the CD variation for 65 nm devices by 80%.

Integrated Maintenance System Trend and A Maintenance Scheduling System Application

Efficient operation and maintenance management for industrial facilities is one of the key issues of social infrastructure systems. This chapter describes some technology trends of integrated maintenance system, especially about monitoring, parameterizing, predicting, and control. Firstly, a trend from Time-Based Maintenance (TBM) to Condition-Based Maintenance (CBM) is discussed. TBM-based agreement between operation owner and maintenance provider is common one, but uncertainty of facility degradation requests CBM application as future service model. Secondly, this chapter describes an optimization method for facility maintenance scheduling, which focuses on the subject that the conflict between maintenance operations and net operations reduces system efficiency.

Business Domain Decision Method for Manufacturing and Service

This paper deals with a study on a method to support the decision on entering business areas in a business model consisting of manufacturing and service. This method divides the business model into business areas by the functions that support a company to decide whether it should enter business areas.This method prioritizes the areas from the aspects of profit opportunities, barriers to entry, and interactions between the areas. This method reduces the time to decide which business area to enter and reduces the number of examination objects to the number of business areas (n) or fewer from the total number of possible combinations (2 to the 2nd power n).To confirm the efficiency of this method, we demonstrated two case studies in the IT services industry. In the first case study, from the perspective of a UPS battery supplier, the number of examination objects is 3 when the number of business areas is 11. And the first case study indicates that a suitable strategy for a battery supplier is to concentrate on the strategy of supplying batteries. In the second case study, from the perspective of an IT service provider, the number of examination objects is 10 when the number of business areas is 11. The second case study indicates that a suitable strategy for an IT service provider is vertical integration from software development to data center operation and hardware manufacturing.Copyright

Product Function: Component Modeling Using Rough Sets

The acceleration of the product development cycle continues to be a significant challenge for manufacturing firms around the world. The misunderstanding of important relationships between product functions and components leads the delay of product development. The present paper describes an identification method of the relationships between product functions and components at the early stage of product development. The proposed product function-component modeling method using rough sets theory extracts the characteristic relationships between product functions and components from a small amount of the qualitative and linguistically-expressed knowledge data. The advantage of using the rough sets is that the combination of necessary and possible sets (lower and upper approximations) represents the vague knowledge. The present paper describes an example of a conventional cutting process with 6 manufacturing parameters that this method contributes to the identification of cutting mechanism from a small amount of sampling data (7% of whole event) compared to the conventional statistical modeling method.Copyright

Probabilistic Evaluation of Product Development Task Planning Using Worth Flow Analysis

The acceleration of the product development cycle continues to be a significant challenge for manufacturing firms around the world. The present paper describes a task planning method that takes the uncertain relationships among the product components into consideration in order to reduce the development time in large-scale and complicated product development with uncertainty at the early stage of product development. We developed a probabilistic worth flow analysis to evaluate each product component for task prioritization with an uncertain relationship among product components. The method calculates the probabilistic distribution of worth flow of each product component based on the probabilistic relationship among product components with the Monte Carlo simulation and determines the development sequence of each component so as to minimize the possibility with the highest feedback information across the task groups. The present paper describes an example of a generic hair drier with a simple mechanical structure developed using the proposed method in order to reduce the uncertainty of feedback information across the task groups while maintaining the uncertainty within same task groups in case the uncertainty has an asymmetric distribution.Copyright

Product Development Task Planning Using Worth Flow Analysis of Interacting Components

The acceleration of product development cycle continues to be a significant challenge for manufacturing firms around the world. This paper describes a task planning method for minimizing trial and error to reduce the development time in large-scale and complicated product development at the early stage of product development. The proposed method matches the group of product components according to geometry and determines the development sequence of each component to minimize the amount of feedback information across task groups. The method applies as evaluation index for task prioritization the product-sum of engineering interaction among components and worth of each component, the “worth flow.” The paper shows with an example of the generic hair drier with simple mechanical structure that this method contributes to the reduction of the size of task group by 22% and amount of information required for setting the interface links by 65% compared to the conventional planning methods.Copyright