INTRODUCTION

In 2019, domestic demand in the Peruvian textile sector contracted by 9%, while exports contracted by 6% (^{Saavedra, 2019}), largely as a result of foreign import prices, which reduces the income of the national textile sector and leads to the closure of several companies. To cope with this situation, companies in the sector are incorporating greater added value to their products, to have tools other than price to compete in the market. ^{The Sociedad Nacional de Industrias (2019}) reports that, although a 3.3% growth in national textile production was expected for 2019, textile exports continue to be low with approximately 2 billion US dollars, compared to the agricultural export sector, which exceeds 7 billion US dollars. Representatives of this institution stated that, in order to solve this problem, it is necessary to work towards increasing the productivity of companies and making them sustainable, so as to reduce costs and wastage, save energy and run more efficient businesses.

The company under study is located in the Constitutional Province of Callao and specializes in the manufacture of firefighter suits made according to existing models in the international market, using inputs imported in their entirety directly from the United States; its products are UL certified and comply with NFPA 1971 and protection standards for the assembly of firefighter suits against structural fire. In view of the current conditions, the company under study found it necessary to improve its processes and reduce its deficiencies to increase productivity. Accordingly, the objective of this article is to propose and implement a management model based on Lean Manufacturing tools, designed to improve the company’s manufacturing process in terms of quality, standardization and time. In addition, our goal is to ensure that this model will not only meet its objective in the company under study, but can also be implemented in other companies in the sector with similar results. The research hypothesis is that there is a significant difference in productivity after the implementation of the continuous improvement model, and that the paradigm obtained is suitable for implementation in similar contexts.

This paper is intended to be useful in the research field as a guide for further studies on process continuous improvement. It is also intended to be useful for those companies in the industrial sector that require a Lean Manufacturing model to improve their processes. The relevance and novelty of this research work lies in the elaboration of a model consisting of a group of tools selected to address the main problems of the garment industry: operator skill level, machine maintenance, lack of order and cleanliness, and high cycle time, which leads to an increase in productivity. Therefore, the research contributes to further development of the topic, since the management model designed constitutes a new alternative for the garment industry in terms of continuous improvement, as well as in the research sector.

The importance of Lean Manufacturing in the management of operational processes of companies is demonstrated in research such as those of ^{Vargas, Muratalla and Jimenez (2018}), who in their article Sistemas de producción competitivos mediante la implementación de la herramienta Lean Manufacturing data on cases of improvement of different production systems. The main problems identified were costs, area used, inventories, quality and lead time, and the result evidenced that 60% of the activities of the companies under study did not add full value; however, this percentage was reduced using the continuous improvement tools. Likewise, González, Marulanda and Echeverry (2018), in their article entitled Diagnóstico para la implementación de las herramientas Lean Manufacturing, desde la estrategia de operaciones en algunas empresas del sector textil confección de Colombia: reporte de caso, demonstrated that in all areas of the textile company under studied where Lean Manufacturing was applied, competitive advantages arise that allow its continuous improvement. Also noteworthy are ^{Curillo et al. (2018}) who in their article entitled Aplicación de herramientas de manufactura esbelta en la empresa textil Anitex, Atuntaqui, Ecuador studied and implemented continuous improvement tools such as Value Stream Mapping, 5S, TPM, SMED, Just-in-Time and Kanban in a textile company. Each of the tools led to improvements; for instance, the application of 5S improved performance in the cutting area by 47%.

METHODOLOGY

A quantitative approach was used in this research. According to ^{Hernández, Fernández and Baptista (2014}), a quantitative approach refers to a set of sequential and evidence-based processes, where the researcher is objective and conducts an organized study based on numerical measurement and statistics and follows a predictable and structured pattern, with which the theories and phenomena investigated can be demonstrated through deductive reasoning from which hypotheses are obtained. The latter are tested to achieve the objective proposed, under an explanatory research level. The research design was pre-experimental. According to Hernández et al. (2014), the pre-experimental design is a study in which one or more independent variables are manipulated, after which the consequences that this action exerts on the dependent variables are analyzed. The unit of analysis was the Production area of the clothing company, from where relevant information concerning the company was obtained. The processes and existing defects in the production of flame-resistant shirts of the clothing company were analyzed, each of which was subjected to evaluation.

The study population was composed of 180 flame resistant shirts equivalent to one month of production. The sample represented 100% of the production of one day, which was randomly selected according to probability sampling. The sample was selected with a margin of error (5%) and a confidence level of 95%, obtaining a total of six units. Data collection was conducted as follows:

All data were classified, recorded and coded using different formats for quality tools, some of them provided by the Lean Six Sigma Institute, which are certified as suitable for use in continuous improvement projects. In addition, descriptive and inferential statistics were used to test the validity of the hypotheses and to obtain conclusions by means of Minitab software and, subsequently, the information was interpreted.

RESULTS

Work hours at the company under study are from 8:00 a.m. to 5:30 p.m., with a lunch break from 1:00 p.m. to 2:00 p.m. This means that there are 8 and a half available working hours (8.5 hours/day) 5 days a week. That is, a total of 42.5 working hours per week.

The daily demand, based on the data of a production order selected for the study with a total of 6 units of flame-resistant shirts, was estimated using the following formula:

Por tanto, la productividad se puede calcular como sigue a continuación:

Calculations were then made by area in order to draw up the company’s current Value Stream Map, which is shown in Figure 2.

The problems that aroused during the production of a production order were quantified according to the number of occurrences. This allowed identifying those that needed to be addressed first in the continuous improvement plan. To this end, a Pareto chart was created, as shown in Figure 3, where it can be seen that 80% of the problems are caused by the lack of order and cleanliness, lack of operator training, unnecessary movements and displacements, and machine failures.

Therefore, in collaboration with the Management area, the Engineering area and the personnel of the Production area, improvement tools were implemented to address each of the aforementioned problems, as shown in Table 1.

After conducting an internal audit of the problems found in the production area, a 5S value of 42% was found, as shown in Figure 4.

Two one-hour training session were held during the morning working hours (when workers are most productive) to implement the 5S. The activities undertaken for the implementation of the 5S in the aforementioned company according to the established schedule are shown in Table 2.

After completion of the final audit at the end of the 5S implementation schedule, a 5S value of 89% was obtained, which is shown in Figure 5.

Subsequently, instruction sheets were prepared for the less efficient operators. An instruction sheet was prepared for each of the machines used in the shirt manufacturing process, showing the beginning, end and process of each operation. These were placed in visible places and explained to the operators. An example of these instruction sheet is shown in Figure 6, which illustrates the one made for the 2-needle lockstitch machine.

To assess operator skill level, a table containing the standard times was drawn up from the time study, which will be shown in the following section. In this table, the machine operations are classified into three levels: A, B and C. As the difficulty of each level increases according to the time it takes the operator to perform the operation, we assigned a weight to each operation based on the evaluation of the workshop manager.

Similarly, a list of skill levels (see Table 3) was drawn up in collaboration with the workshop manager, who trained the operators in the production of the shirts, to calculate the skill level acquired during training.

Based on these data, the skill level percentage of each of the operators in each operation was calculated using the following formula:

As there are a total of 36 machine operations, the calculations were summarized to show the initial and final skill level of each operator. Note that the “initial” is the training given by the workshop manager, and the “final” is the one after the implementation of the instruction sheets.

The following average initial skill level values were obtained for operator Yorely S.:

% Skill level = (62%×10+86%×10+⋯+43%×35+59%×35)/875=56%

After the implementation of the instructions, the following average final skill level value was obtained for the operator Yorely S.:

% Skill level = (79%×10+90%×10+⋯+51%×35+62%×35)/875=71%

The following average initial skill level values were obtained for operator Sandy C.:

% Skill level = (45%×10+69%×10+⋯+42%×35+61%×35)/875=47%

After the implementation of the instructions, the following average final skill level value was obtained for the operator Sandy C.:

% Skill level = (58%×10+72%×10+⋯+56%×35+74%×35)/875=68%

Prior to the implementation of the 5S and instruction sheets, and due to the fact that the efficiency of these operators is initially low, a time study was conducted. Very long times were recorded in the manufacturing process, some of which had to be eliminated to estimate the standard time. The valuation factor (VF), equivalent to operator efficiency, is a value assigned by the Engineering staff. The tolerance factor (TF) of each of the operators is calculated using the additional times from the OIT table. Once the calculations were made, it was found that the manufacture of a long sleeve flame-resistant denim shirt for the production order in the study took 64.615 minutes.

Upon implementation of the two aforementioned tools, the standard time for the manufacture of flame resistant shirts with the same characteristics decreased to 50.559 minutes. In addition, a Failure Modes and Effects Analysis was conducted, as shown in Figure 7, in order to determine the main failures that could occur during the manufacturing process. The degree of severity, detection and occurrence of these failures was also measured, and solutions were proposed through the implementation of TPM.

After noticing the presence of critical points in the FEA, the initial OEE was calculated (Figure 8) to gain a rough idea of the overall effectiveness of the equipment.

In view of the severity of the problems presented by each machine, a total productive maintenance plan was implemented for each one of them, an example of which can be seen in Figure 9. It was also necessary to provide daily maintenance to the machines before and during operations and at the end of the shift. Maintenance is performed in a few minutes, in order to inspect the proper functioning of the machine before it starts operating and thus reduce time losses due to machine failure.

.After the implementation of the preventive maintenance plan and daily maintenance, the OEE value was obtained, as shown in Figure 10.

Initially, the OEE was 56%, which is unacceptable according to the metric presented in Table 4. In other words, the equipment was initially quite unproductive, which represented significant economic losses. After the implementation of the preventive maintenance plan and the daily maintenance log, a value of 84% was achieved, which is acceptable according to the metric. Although there are still slight economic losses and competitiveness is slightly low, the OEE is nevertheless higher than it was initially.

Finally, after the implementation of the proposed tools, the future value stream map is drawn up (Figure 11), based on the new time values and the company’s inventory.

At the end of the implementation of the tools, the productivity value was recalculated and the following values were obtained:

Comparing it with the initial value obtained, we have:

Following the experimentation conducted in the company and in view of the positive results obtained, the Lean Manufacturing model for clothing companies was determined. A questionnaire was elaborated to validate it, considering the research variables, and was applied to 8 experts in the field. The results were used for the Aiken’s V test (averages of the value assigned to each item), whose values obtained can be seen in Table 5.

*p < .05

Experts = 8

It is worth mentioning that the last item corresponds to Observations, which is to indicate whether any correction, change or improvement should be made to the model, where it is possible to choose Yes (1) or No (0).

The Aiken’s V coefficient is obtained with the following formula:

Where:

Using the formula above, we obtain:

As 1 > 0.80, the Lean Manufacturing model proposed in this research is valid for its application, with 100% validity by Aiken’s V test.

Upon obtaining these results, the hypothesis testing was performed to confirm the assumption made at the beginning of the research. First, the Saphiro Wilk, for samples smaller than 50, was performed with Minitab 2018 statistical software to determine whether or not the data followed a normal distribution with the following parameters:

p ≤ α, data do not follow a normal distribution

p > α, data follow a normal distribution

Based on the data, a p-value equal to .10 was obtained, which is greater than the significance value of .05; therefore, the data have a normal distribution. Following this demonstration, a Student’s t-test was performed for the acceptance or rejection of the hypotheses, which is presented in Table 6, considering the following premises:

H₀ = There is no significant difference in productivity improvement before and after the implementation of the continuous improvement model.

H₁ = There is a significant difference in productivity improvement before and after the implementation of the continuous improvement model.

Considering that a p-value equal to .347 was obtained, H₀ is rejected and H₁ is accepted. Therefore, there is a significant difference in productivity after the implementation of the continuous improvement model. Consequently, it is understood that the proposed management model has achieved its objective. This result can be extended to other organizations by implementing the model in other environments, where, having demonstrated its validity and functionality, productivity improvement results will be obtained.

DISCUSSION

During the validation of the Lean Manufacturing Model, all the experts interviewed agreed on an acceptable value of Aiken’s V, thus it can be said that this model contributes to the productivity of the companies in the sector. This statement is based on the use of 5S, VSM, implementation of instruction sheets, time studies and TPM. These results are consistent with those of ^{Gonzales and Samán (2020}), who state that the implementation of these tools led to an increase in business productivity from 36% to 80%, doubling the number of items produced per day. Likewise, ^{Quispe and Requena (2019}) demonstrated that the application of these tools improved production times, efficiency and productivity, with which they managed to achieve a 29% increase in the latter indicator. ^{Huaripata and Matos (2017}) found that the implementation of these tools reduces raw material cost overruns by 13%.

CONCLUSIONS