Improving safety of crane-related operations in chemical industry by the support of a real-time computer-aided visual guidance system

Safety is a fundamental issue in all workplaces. The literature shows that cranes belong to the class of the most dangerous equipment in the industrial context and it has also been observed that several accidents are due to a partially or totally obstructed view of the working area for the crane-operator. Thus, a dropped or mishandled load can directly injure workers or causes damages to the surrounding. Accidents could lead to the release of dangerous materials in the chemical process industry and in the intermodal transport. To contribute to the reduction of crane-related accidents due obstructed view, a real-time Visual Guidance System (VGS) has been developed within the SPRINCE project (acronym of Smart PRocess INdustrial CranE). The use of this application, once the related prototype hardware for the image acquisition has been installed on the crane, aims predicting whether a dangerous event (i.e. a potential collision with an obstacle) is going to occur and promptly alerting the crane-operator. Its feedback allows taking corrective actions during the execution of crane-assisted shifting duties. An earlier version of such a system was presented in the conference ESREL 2016, this further contribution shows the results of some tests made for the validation of the VGS in an industrial context. It describes the performance of the prototype and its application when used in a real case-study. Some criticalities have also been found due to the complexity of the surrounding, where the load is going to be lifted, thus further improvements for the system were necessary.