Linguistic analysis for occupational risk assessment communication using the Lpac in an AI environment

Abstract

Risk assessment methodologies results in construction sector are usually reported within a very limited linguistic context of preventive action, which may involve interpretation difficulties. This research identifies the problem of risk assessment methodologies to linguistically convey the results of risk assessment to construction workers. The methodology adapted to construction sites called Level of Preventive Action (Lpac) determines levels of preventive action control related to a spectrum of quantitative results, with data collection based on physical conditions of the environment with sensors in the construction and safety systems; and behavioural conditions and worker emotional state with body sensors, locators and integrating with the BIM approach. The communication of the results must be adapted to the multiple variables that occur in a communicative interaction. The Lpac method proposes a broad spectrum of preventive action control conditions on the physical and human environment that are transmitted to the worker, which allows for different communicative possibilities. In this sense, safety is assessed concerning the different risks involved in work performance, applying control banding concepts. Optimising the physical and human environment in real time is essential through devices that generate a complex human-machine interaction within an artificial intelligence environment in the construction sector.