Controle estatístico da qualidade em um processo de envase da indústria de alimentos

• Ano de defesa: 2020
• Autor(a) principal: Rique Júnior, José Flávio
• Tipo de documento: Dissertação
• Tipo de acesso: Acesso aberto
• Idioma: por
• Instituição de defesa: Universidade Federal da Paraíba; Brasil; Engenharia de Produção; Programa de Pós-Graduação em Engenharia de Produção; UFPB
• Palavras-chave em Português: Controle Estatístico do Processo (CEP); Cartas de Controle; Matriz de Priorização; Análise de Concordância de Atributos; Statistical Process Control (SCP); Control Charts; Prioritization Matrix; Attribute Agreement Analysis; CNPQ::ENGENHARIAS::ENGENHARIA DE PRODUCAO
• Link de acesso: https://repositorio.ufpb.br/jspui/handle/123456789/20739

Statistical Process Control (SCP) is an effective tool in helping to reduce variability and stabilize processes. SCP, with its benefits, can be an important ally to the food industry, which in turn suffers from the high variability of its processes. This high variability is due to several factors, for example, seasonality of the raw material and the high perishability of the products, so, due to this, the rate of defective items in the food industry can be high. The general objective of this work is to reduce the rate of defective sachet items in a food industry, through control charts, capability analysis, prioritization matrix and attribute agreement analysis. For this first, the measurement of the current state of the process was carried out, through the development of a method of implantation of the Statistical Process Control in the sector of filling of fruit pulp sachets, structured according to Phase I and Phase II of operations of control charts. This method was subdivided into 6 steps, problem statement, data collection plan and calculation of the initial control limits, stability analysis, capability analysis, online monitoring, and finally, the decision to end the process monitoring. Soon, data related to approximately 2 months of production were collected and analyzed. In Phase I, of control limit calculations, and stability and capability analysis, Special Causes of Variation (SC) were found, and corrections and preventions against recurrence were made. The Defects Per Million (DPM) index was obtained, corresponding to 21170 defective products for every 1 million produced, equivalent to a 2σ level process. In Phase II, the online monitoring phase, with the effective use of the control charts, and after correcting and preventing Special Causes (SC), the stability of the process was achieved. In view of the defect rate corresponding to 21170 defective sachets for every 1 million produced, a structured search for the causes of this high rate began, with meetings held with the specialists involved in the process to search for these possible causes. In the first meeting with the experts, a brainstorming was carried out, where each alleged cause was classified according to the 6M's of the Ishikawa diagram, with 36 possible causes being listed. Then, each specialist allocated each cause according to the rankings of a prioritization matrix. After allocations, the statistical tool, attribute agreement analysis was used to validate the classifications against a pre-established standard. Then, a filter was carried out based on the analysis of attribute agreement, through criteria of prioritization and elimination of causes, until 12 possible causes remain among the initial 36. With the 12 causes, the action plan was established and short and medium term goals were established for their solutions. With the fulfillment of the solutions for the causes established in the short term, it was possible to carry out the validation of the method by repeating the entire initial procedure, calculating new control limits and a new DPM equivalent to 14978 defective items for every 1 million produced. Compared to the initial one, after the implementation of the method, there was a 30% reduction in defective items.