This paper addresses safety challenges due to errors and misbehaviours of the operators in complex production systems when malfunctioning and failures must be manually solved by the operators or when defeating of the machine safety devices is carried out. To this purpose, in this paper a novel integrated supervision system is proposed. The system uses various technologies, including RFID tags capable of detecting both personnel and object positions to gather information from the working area, computer vision (CV) technology to observe and analyse the behaviour of both the operator and the machine and Artificial Intelligence (AI) module. The comprehensive dataset which is collected by the RFID and CV systems together with the knowledge of the current operative state of the assembly is transmitted to the AI module, which, following a period of thorough training, is expected to identify potential risks to personnel. Then, the AI module will trigger safety functions, predetermined by the machine manufacturer, in response to deviations from established safety standards. Output signals may include warning notifications to the workers on portable mobile devices, advising against proceeding with production. The whole integrated system is completed by a front-end software for its management by the responsible of the assembly. To demonstrate the efficacy of this novel concept, we plan to develop a prototype connected to a production system comprising a lathe and a collaborative robot (cobot). This production system will integrate data pertaining to work phases, machine and cobot status. In this collaborative process, the cobot and lathe work together to automate the turning process, adhering to stringent safety conditions. To validate the effectiveness of the supervision system, rigorous testing will be conducted in common risk scenarios associated with the turning activity of the prototype. The main goal of this innovative device is to increase the workplace safety reducing the effects of human errors and misbehaviours.
Development Of Novel Integrated Smart System Based On Artificial Intelligence For Management Of Operator Safety In Production Processes
Roberto Gabbrielli;Martina Olivieri;Francesco Di Paco;Leonardo Marrazzini;Marco Frosolini;Marcello Braglia;Claudio Gennaro;Francesco Marcelloni;Paolo Nepa;
2024-01-01
Abstract
This paper addresses safety challenges due to errors and misbehaviours of the operators in complex production systems when malfunctioning and failures must be manually solved by the operators or when defeating of the machine safety devices is carried out. To this purpose, in this paper a novel integrated supervision system is proposed. The system uses various technologies, including RFID tags capable of detecting both personnel and object positions to gather information from the working area, computer vision (CV) technology to observe and analyse the behaviour of both the operator and the machine and Artificial Intelligence (AI) module. The comprehensive dataset which is collected by the RFID and CV systems together with the knowledge of the current operative state of the assembly is transmitted to the AI module, which, following a period of thorough training, is expected to identify potential risks to personnel. Then, the AI module will trigger safety functions, predetermined by the machine manufacturer, in response to deviations from established safety standards. Output signals may include warning notifications to the workers on portable mobile devices, advising against proceeding with production. The whole integrated system is completed by a front-end software for its management by the responsible of the assembly. To demonstrate the efficacy of this novel concept, we plan to develop a prototype connected to a production system comprising a lathe and a collaborative robot (cobot). This production system will integrate data pertaining to work phases, machine and cobot status. In this collaborative process, the cobot and lathe work together to automate the turning process, adhering to stringent safety conditions. To validate the effectiveness of the supervision system, rigorous testing will be conducted in common risk scenarios associated with the turning activity of the prototype. The main goal of this innovative device is to increase the workplace safety reducing the effects of human errors and misbehaviours.I documenti in IRIS sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.