Use of Inventory Control Theory and Multi-Objective Optimization to Model Work-Rest Scheduling

• Main Author: Sarker, Pramiti
• Format: Dissertation
• Language: English
• Published: ProQuest Dissertations & Theses 01.01.2022
• Subjects: Engineering; Health care management; Industrial engineering
• ISBN: 9798368430461

The research aims to study the effect of varied work-rest scheduling strategies on muscle fatigue development and worker performance. An inventory control theory modeling approach was used to model the level of muscle fatigue under different break schedule strategies and a multi-criteria optimization formulation was employed to simultaneously consider the competing interests of reducing muscle fatigue and increasing worker productivity to determine an optimum number of breaks. Several published preliminary studies prepared the foundation for this modeling approach. The first study (Chapter 2) was conducted to study the effect of breaks on the slope of median frequency of EMG signal (a standard measure of muscle fatigue development). Downward shifts in the median frequency (MDF) of the power spectrum of an electromyographic signal is often used to assess muscular fatigue. How the change in the MDF is affected by repetitive bouts of exertions–with intervening rest breaks- is not well understood. It was hypothesized that repetitive bouts of a fatiguing, isometric exertion, separated by periods of rest, would have cumulative effect (across bouts) on the slope of the decline of the median frequency, with an expectation of an increasing rate of decline in subsequent bouts. To test this hypothesis, 24 participants performed four bouts of an isometric (15% MVC) elbow flexion exertion. Each exertion lasted for four minutes and then a 15-minute break was provided between bouts. Surface electromyography was used to capture the activity of the biceps brachii at twenty-second intervals during the exertions. The median frequency of each of these five-second collection periods was calculated, as was the slope across the four-minute bout. The results showed that there were no statistically significant differences in the rate of decline in the median frequency across bouts. The study helped to interpret median frequency as a measure of muscular fatigue in an ergonomic intervention. The results showed no effect of break on median frequency slope, indicating that break helped them in developing no long-term fatigue/break helped them to recover and imposed the importance of finding the right ratio of work to break. Based on these results the next study (Chapter 3) was conducted. This study focused on neck muscle fatigue replicating the real-life work-environment where prolonged static neck posture was adopted to perform a task. In real-life, sustained non-neutral postures of the head/neck are related to transient neck discomfort and longer-term disorders of the neck. From the previous study (Chapter 2), we obtained the knowledge that periodic breaks did not increase the slope of median frequency decrease, that is, the fatigue was not carried over to second bout and subjects were recovering. However, the ideal frequency and duration of breaks was yet to be determined. Therefore, the next study (Chapter 3) aimed to quantify the effects of three work-rest strategies on fatigue development. Participants maintained a 45-degree neck flexion posture for a total of 60 minutes and were provided three minutes of rest distributed in different ways throughout the experiment [LONG (one, three-minute break), MEDIUM (two, 1.5-minute breaks), or SHORT (five, 36-second breaks)]. Surface electromyography data were collected bilaterally from the neck extensors and trapezius. Subjective discomfort/fatigue ratings were also gathered. Results of the analysis of the EMG data revealed that the SHORT condition did not show increased EMG activity, while LONG [21%] and MEDIUM [10%] did (p<0.05), providing objective data supporting the guidance of short, frequent breaks to alleviate muscular fatigue. The study helped us to find the similarity between trend of muscle fatigue-recovery with production and delivery trend in inventory control theory. The study also shed light on the fact that frequent work-rest bouts reduced muscle fatigue, but it also raised concerns regarding the impact on productivity of the workers. The optimum number of breaks based on both muscle fatigue and performance are yet to be determined. The subjects in this previous study (Chapter 3) performed a task a game-based task and performance was not quantified. The fourth chapter was designed to determine the optimum number of breaks that can minimize fatigue and maximize performance simultaneously. An optimum number of breaks would reduce the negative impact of muscle fatigue without impacting performance. To develop this model, an analogy between inventory control model and fatigue development model was created. Finding optimum production quantity in an inventory control model helped to create the idea of finding optimum number of breaks in this work-rest scheduling model. The work with a laparoscopic simulator was chosen as a work task. To develop this model, 17 subjects were asked to come to practice with the laparoscopic simulator and learn how to operate laparoscopic instruments. When they were proficient, they participated in five experimental sessions (each consisting of 23 minutes of work with a laparoscopic simulator) for five different conditions. Five sessions have five different work-rest schedules as conditions. These conditions aimed to observe the effect of a work-rest schedule on muscle fatigue development and performance. An equation of fatigue vs the number of breaks and the equation of performance vs the number of breaks were quantified. The resulting equations showed conflicting relations with the number of breaks. If the number of breaks was increased, muscle fatigue decreased. On the other hand, performance increased with increasing number of breaks and then started decreasing. These two conflicting equations developed based on inventory control theory model, formed a multi-objective problem that returned the optimum number of breaks. Multi-objective optimization problem can be solved in different ways. Weighted average method was applied to solve this problem. Each optimum number of breaks for different weights (between 0 to 1) is a pareto optimal solution. In this study, Entropy Weight Method (EWM) was applied to determine the value of weight without the influence of personal opinion. The method found a weight of 0.57 which was applied to the objective function for muscle fatigue and a weight of 0.43 was applied to the performance. For a weight of 0.57, the model output suggested 11.1 breaks during 23 minutes of high intensity work. The model also expanded to explore the effects of inter-individual variability on these predictions by finding the number of breaks for different percentile value of muscle fatigue and performance. The result suggested that with a very low fatigue profile and performance score, no break is needed. When high performance is expected from a high fatigue profile person, then number of breaks should be around 14 breaks. After 14 breaks, performance started to decrease significantly, and number of breaks should not exceed 14, if both performance and fatigue are expected to be optimized during a 23-minute period.