Manufacturing sites play host to a wide range of operational environments and equipment that may pose safety risks to employees. Consequently, the issue of how to protect personnel from workplace dangers has long been a primary concern for government agencies, corporates and laborers. According to statistics from the U.S. National Safety Council, the total cost due to labor injuries across industries in the U.S. was US$171 billion in 2019. Statistics from the U.S. Department of Labor show that in 2020, the case count of work-related nonfatal injuries and illnesses in the private sector totaled 2.654 million; among them, 373,000 were in the manufacturing industry, accounting for 14.06% of the total.
Injuries to manufacturing personnel not only affect the physical/mental health of workers and their families, but also leads to additional operating costs and losses for companies as a result of injury compensation settlements and reduced output. Hence, it requires a constant effort from both corporates and employees to ensure manufacturing personnel’s safety and to prevent workplace injuries.
According to the annual report of labor inspection statistics issued by the Occupational Safety and Health Administration of the Ministry of Labor, the number of reported accidents involving personnel in the manufacturing sector resulting in disability or injury totaled 5,147 in Taiwan in 2020. Statistics from the Ministry of Labor indicate that the largest categories of personnel injuries were being crushed or caught up in machinery (21.97%), falls at the manufacturing sites (18.22%), cuts, scrapes, and scratches (16.22%) and knocks, inappropriate actions, and burns.
According to a survey by the U.S. personal protective equipment maker Kimberly-Clark Professional, one of the primary reasons for occupational injuries is employees’ non-adherence with the requirements for wearing protective equipment. Nearly 90% of safety officers have experience of dealing with workers not wearing protective equipment as required. Almost 30% of safety officers said it was a frequent occurrence. Therefore, it is necessary to adopt a more thorough, real-time and continuous method of inspection and monitoring than by just using safety officers to ensure the safety of manufacturing personnel and their compliance with safety SOPs (Standard Operating Procedures).
The use of artificial intelligence (AI) to enhance the safety of manufacturing personnel offers the advantages of continuous monitoring and real-time alerts, as well as the early detection of potential risks. Despite the initial investment in hardware and software, it remains a worthwhile solution for manufacturers as it prevents injuries and reduces losses due to production suspension, insurance, and compensation as a result of industrial safety issues.
Whilst many manufacturing sectors have been introducing automated production systems, many procedures still rely on manual operations due to considerations of production flexibility and costs. In order to maintain productivity and product quality, an issue of concern to the industry is not only necessary to formulate SOPs according to manufacturing processes, but also to ensure that personnel adhere to these SOPs and continued efforts are made to enhance production efficacy with improvements in operational methods.
Machine vision and AI can help to continuously monitor whether personnel follow the SOPs and thus avoid the problems resulting from inadequate human supervision. In addition, the image analysis of footage of personnel performing operational procedures can help industry players to identify directions and opportunities for operational effectiveness improvement. Therefore, these applications will be increasingly valued and adopted by industries.
The combination of machine vision and AI will enable the manufacturing industry to develop innovative solutions to reduce the work safety risks to personnel, provide better protection and continue to improve operational efficacy. Therefore, AI-related technologies such as smart cameras; the tagging, labeling and identification of manufacturing environments, machinery, objects, and personnel; and analytics of human movements, behavior and operations are all going to become important technologies and products that require ongoing R&D investments. Many research organizations and companies, both domestic and foreign, have been developing related technologies, applications and solutions.
In Japan, Fujitsu launched a new AI technology in January 2021, for the identification of complex human movements from multiple joint locations via deep learning. The German Research Centre for Artificial Intelligence (DFKI, Deutsches Forschungszentrum für Künstliche Intelligenz) is studying how AI can be used to identify any physical discomfort manufacturing personnel might be experiencing. Cisco has integrated edge computing devices and AI technology to identify whether personnel are wearing protective equipment according to requirements.
Manufacturing sites are relatively complex and changeable environments with many inherent safety risks. The protection of personnel safety will remain the primary goal for corporates, managers and laborers, whether it is on processing and assembly lines that are highly dependent on human labor or at automated facilities with a variety of machinery. The use of AI to enhance the safety of manufacturing personnel offers the advantages of continuous monitoring and real-time alerts as well as the early detection of potential risks. Despite the initial investment in hardware and software, it remains a worthwhile solution for manufacturers as it prevents injuries and reduces the loss due to production suspension, insurance, and compensation associated with industrial safety issues.