This Is AuburnElectronic Theses and Dissertations

The Effects of Lean Manufacturing on Occupational Safety and Ergonomics




Brawner, Joel

Type of Degree

PhD Dissertation


Industrial and Systems Engineering


The association between Lean Manufacturing (LM) and occupational safety and ergonomics (S/E) is a widely disputed topic in scientific literature. This study attempts to add clarity to this controversial relationship through literature review and experimental analysis. From the literature review, one hundred and one studies containing one hundred and seventy LM-related safety/ergonomic outcomes were identified. Thirty-seven outcomes pertained to the use of Just-in-Time (JIT) production, which were overwhelmingly negative in nature. Conversely, the twenty-six studies pertaining to 5S contained almost exclusively positive outcomes. No outcomes from the adoption or use of Lean Culture were found. These review findings suggest that individual LM methods, especially JIT and 5S, uniquely contribute to the safety/ergonomic outcomes attributed to LM, while the effects from Lean Culture, the principle that emphasizes respect for the worker, remain unknown. Based on the review findings, a method was designed to test the effect of LM principle adoption levels on S/E. Surveyed employee perception of LM adoption levels in the principles of Standardization and Stability, Built-in-Quality, JIT, and Lean Culture were used to define the independent variables; based on a 7-point Likert scale. The dependent variables were made up of safety incident rates from the Occupational Safety and Health Administration (OSHA), normalized to industry standard rates from the Bureau of Labor Statistics (BLS), for each studied manufacturing facility. These dependent variables took the form of normalized total case rate (TCR); rate of cases requiring days away, restricted, or transferred (DART); and rate of cases requiring days away from work due to injury or illness (DAFWII). Regression analysis and structural equation modeling (SEM) methods were the selected methods to study the relationships. Survey invitations were sent to 361 American manufacturing facilities, and after removal of invalid responses and extreme outliers from the received responses, 271 valid responses from 15 facilities remained. Dependent variables from the associated facility were added to each survey response to complete the data set. An exploratory analysis of this data resulted in some hints of association between both Culture and JIT adoption levels on incidence rates, but none pertaining to Standardization and Stability. Results from multiple regression analysis confirmed these indications, finding significant effect coefficients pertaining to both JIT and Culture adoption levels, yet still no significant results pertaining to Standardization and Stability, the principle containing 5S. These regression results were successfully corroborated through SEM, robust regression, and non-parametric tests; due to the non-normal distributions and Likert data contained in the data set. The analysis found that higher employee-perceived Lean Culture adoption was associated with lower TCR and DART rates, while higher perceived JIT adoption was associated with higher incidence rates. The effect coefficients indicate that a 1-point Likert-scale increase in perceived JIT adoption level was associated with 0.445 more total cases and 0.346 more DART cases per 100 workers per year, while Culture adoption indicated 0.518 fewer total cases and 0.373 fewer DART cases. Adjusted R2 values from this analysis indicated that this JIT/Culture multi-regression model explained 5% of the variance in total case incident rate and 6% of the variance in DART rate. The results from this study suggest that LM implementation does indeed have a small influence on S/E, and that influence is beneficial so long as Culture adoption occurs at an equal or higher rate than that of JIT. However, JIT adoption with lagging or absent Culture adoption can result in a negative S/E effect in an LM implementation.