Analysing an industrial safety process through process mining: A case study

Abstract

The application of data analytics has delivered significant value in a broad range of industries, but has frequently failed to bring about operational efficiencies or process safety improvements in asset-intensive sectors despite a profound increase in the volume of digital information that is stored by companies operating in this domain. Process safety management is a major concern of asset-intensive organisations, as the consequences of incidents can be catastrophic. In order to prevent incidents from occurring, organisations enforce prescriptive safety procedures frequently known as “Permit to Work” systems. These procedures can be thought of as a workflow consisting of a set of well-defined steps that have to be performed in a certain order to minimise safety risks. In this paper, we apply analytical tools from the area of process mining to shed light on how safety processes are actually executed. Process mining constitutes a relatively recent area of research at the intersection of data analytics and business process management (BPM) that is concerned with extracting insights from event logs that record process executions. We report the results of a process mining case study based on a Permit to Work system in an asset-intensive organisation in Australia.