THE COMPETITIVE SEMICONDUCTOR MANUFACTURING HUMAN RESOURCES PROJECT:

Second Interim Report
CSM-32
Clair Brown, Editor

7. Skills and Work Tasks
Melissa M. Appleyard

7.1 Introduction
7.2 Teams
7.3 Equipment Maintenance
7.4 Statistical Process Control
AP Appendix
FIG Figures 1-16

7.1 Introduction

Rapid and effective problem-solving is a critical activity in the semiconductor industry. By examining both the workings of teams and the integration of problem-solving activities into the job tasks of operators, technicians and engineers, this chapter presents suggestive evidence of management practices that lead to successful problem-solving in a semiconductor fab.

Section 7.2 discusses the composition and coordination of teams in the fifteen fabs in the HR subsample. Teamwork is commonplace in the majority of fabs in our sample, and teams at different fabs share many characteristics. The structuring of work tasks and the participation of workers in specific activities seem to have more of an impact on performance than the characteristics of teams.

Equipment maintenance and statistical process control (SPC) activities analyzed in Sections 7.3 and 7.4, respectively, exhibit a greater degree of variation in human resources practices across the fifteen fabs than do the data on teams. Section 7.3 describes how the fabs allocate equipment maintenance tasks across three primary occupations: operator, technician, and equipment engineer. Similarly, Section 7.4 focuses on the allocation of SPC activities across the operator, technician and process engineer occupations.

Both equipment maintenance and SPC are central to effective problem-solving in a fab. Given this, we hypothesize: Fabs that engage in technical tasks most intensively should exhibit the highest performance. The correlations between performance measures and total fab SPC in Section 7.4 support this hypothesis, although the correlations involving equipment maintenance found in Section 7.3 do not. We also hypothesize that fabs that include their "front-line of defense," namely their operators and technicians, in equipment maintenance and SPC achieve a higher level of performance. With respect to the operator occupation, evidence in support of this hypothesis is found in both Sections 7.3 and 7.4. The data in Section 7.4 also highlight that process engineers play a vital role in successful SPC activities. A fab's need to solve problems quickly and permanently requires operators and technicians to identify problems immediately and then work with engineers to uncover root causes and implement lasting solutions. The actual participation of workers in these problem-solving and maintenance activities is more important than their membership in teams per se.

7.2 Teams

By examining the workings of teams in the fabs in our sample, we wanted to document options for structuring group problem-solving. Fourteen of the fifteen fabs in our sample reported having at least one of the following types of teams: Quality Improvement Teams/Quality Circles (QITs/QCs), Self-Directed Work Teams (SDWTs), or Cross-Functional Teams (CFTs). Table 7-1 lists the fabs in our sample and their types of teams. Although many fabs have instituted group problem-solving activities, our interviews with teams during site visits taught us that their level of effectiveness varies greatly.

Table 7-1. Presence of Teams

Table 7-2 contains the definition for each type of team found in the questionnaire. The definitions reveal that the teams differ along two primary dimensions: their area of focus and their level of autonomy. As for their focus, QITs/QCs and SDWTs concentrate on problems in their immediate work area, whereas CFTs draw members from a number of work areas. Of the three types of teams, SDWTs generally have the greatest level of autonomy. Across the fabs in our sample, QITs/QCs and CFTs are the most pervasive (11 fabs reported having such teams), with the median number of such teams being 13 and 9 respectively.

Table 7-2. Team Definitions

Given the nature of the teams, we predicted that SDWTs would have the greatest number of meetings a week, have compulsory membership, exist for more than a single project, and have the greatest level of autonomy for project selection and expenditures. As Table 7-3 shows, the data support our predictions to some degree. However, managers play a larger role in SDWTs (as they do in all types of teams) than we anticipated.

Table 7-3. Characteristics of Teams

According to the data presented in Table 7-3, CFTs and QITs exhibit very similar characteristics. The fabs in our sample emphasize both problem-solving across multiple work areas and quality improvement activities in a single work area to a similar degree. The importance placed on cross-functional problem-solving reflects an interesting feature of the semiconductor industry: The complicated interplay of processing steps requires that workers in different equipment areas communicate regularly.

Across the fabs in our sample, the three types of teams are very similar in terms of: size, meeting length, and use of structured problem-solving techniques. For all three types, the average number of members is approximately 10, with the average for SDWTs being the highest (13.3), and the average for CFTs being the lowest (7.5). An Asian fab reported having the largest teams, QITs/QCs with 30 members, while two other Asian fabs reported having the smallest teams, CFTs with 4 members. The meetings for all the teams in the sample last 1-2 hours. Only two fabs reported that their teams do not use formal problem-solving techniques: one Asian fab with QITs/QCs and one U.S. fab with SDWTs.

Table 7-4 presents the share of team membership across the three primary occupations. Relative to operators and engineers, technicians have the smallest average membership share in QITs/QCs and CFTs. On average, operators constitute over 60% of the team members for both QITs and SDWTs which is similar to their share of total headcount discussed in Chapter 3. Engineers constitute the highest average share of CFT membership, over 40%, which is triple their share of total headcount at many fabs in our sample. This high level of participation by engineers in CFTs reflects the technical challenges faced when problems span multiple work areas. Such problems require engineering expertise and leadership for resolution, since successful semiconductor manufacturing processes rely on the robust integration of process steps. The average percentage of technicians does not exceed 16% for any of the teams, which falls just below their average share of headcount.

Table 7-4. Team Membership by Occupation

Go to Next Section of this Chapter
Go to Table of Contents for this Chapter
Go to Table of Contents for the CSM-HR Interim Report