Constructivist multi-criteria model to support the management of occupational accident risks in civil construction industry.

Abstract

Civil construction is one of the industrial sectors with continuous growth globally, particularly in Brazil in the last 50 years. Unfortunately, it is also one of the productive segments with the highest incidence of accidents, which overshadows its merits as a driver of economic growth and job creation. The damage to workers' health caused by accidents at work results from the presence of work environment risk factors. Therefore, this study aims to manage these risk factors for the civil construction industry. The work is original with respect to building a model to support risk management in civil construction for a specific and relevant context. It is ensured by presenting an unprecedented approach to the sector that incorporates information not considered by classic generic approaches. This research, thereby, seeks to build a model to support the risk management of accidents in the workplace in the prefabricated concrete construction industry. It is a case study with a constructivist approach and an exploratory and descriptive character, incorporating the Multicriteria Methodology for Decision Aiding-Constructivist (MCDA-C). The main findings include (i) identifying the strategic objectives: occupational safety policy, work environment, machines and equipment, condition of materials, procedures and methods, and skills, which were operationalized via 58 criteria; (ii) examining the scales of the criteria such as the performance profile of the current situation and the goal, highlighting the vulnerabilities and potentials; (iii) proposing improvement actions for the vulnerabilities, thus supporting risk management in the organization. Among the contributions, managers and professionals in the field contribute to the possibility of using an instrument customized to the context and legitimate to their concerns and values stands out. Furthermore, the contributions of researchers include the challenge of improving their generic models with the knowledge of personalized models.