Industrial development of car disassembly - ergonomics and system performance

Sammanfattning: A new EU directive on used vehicles has recently been introduced. It demands that for every scrapped car, at least 85% by weight must be recycled by the year 2006 and 95% by 2015. The current level is about 80%. The car disassembly industry was chosen as the study object of this thesis. Due to the increased demands on recycling, the disassembly industry may undergo comprehensive rationalizations and expand into a modern mass-type production. Rationalized car disassembly systems may be an alternative to shredding processes. The general aim of this thesis is to study the connections between expected rationalizations and ergonomics based on the case of the car disassembly industry. The studies in this thesis aim to collect empirical data on present production system performance and ergonomics (physical workloads) and possible future development. Car dismantlers, manufacturers and authority stakeholders described current disassembly systems as ?craft-type?, i.e. containing a rich variety of tasks and considerable periods of set-up time. The dismantling companies reported good profitability. Expectations about production systems in the future were that the amount of non-profit work per car would increase and more parts and materials than at present will have to be dismantled. These materials lack market value today. For production to stay profitable in spite of increased time in non-value-adding tasks, comprehensive rationalizations were anticipated. Design for Disassembly/Recycling was not fully applied at the manufacturing. However, the need for this strategy was expressed by both dismantlers and manufacturers, as was the need of cooperation between these two groups. The ?craft-type? disassembly workers reported a high physical workload for the arm and for the low back. Pain levels were highest for the low back, with 29.5% of operators reporting pain to occur ?often? during the last 12 months. Disassembly workers had higher pain scores than a general male population in Sweden. The psychosocial working conditions, i.e. demands, influence and social support were lower than the corresponding data for the Danish working population. Technical measurements of physical workloads and time consumption in the current ?craft-type? Swedish car disassembly were made. Disassembly work implied high circulatory loads, much walking and high peak low back loads. Value-adding, ?direct?, work comprised only 30% of the total working time, and implied more awkward postures and higher movement velocities for the head, arm, upper back and wrist than non-value adding tasks, as well as less time in rest. The physical workloads and operators? utilization of time were also assessed in a serial-flow ?industrialized? production system for car disassembly in the Netherlands. Time proportions of direct work as well as body postures were similar in the serial-flow car disassembly and in the Swedish craft-type. Peak low back load tended to be smaller in the serial-flow system, while the upper limb movement velocities appeared to be higher in this system. The serial-flow disassembly system showed production deficits due to factors such as system losses, worker inexperience and teamwork deficits. A novel combination of flow and biomechanical simulation was presented in order to assess the physical loading consequences of alternative system configurations. A smaller variation in cycle times implied higher output in number of cars per week and larger operator cumulative loading on the low back. Reducing cycle times, on the other hand, resulted in higher output without significant change in utilization rates and thus unchanged cumulative load. Combined human and flow simulations may allow an integrated consideration of productivity and human factors in the early system development. Reliability of the video-based tool for work task analysis was assessed. Task analysis was one of the core methods used in this research, and it may be more frequently used in ?rationalization? studies. In general, there was good agreement between observers both on overall task proportions and on the mean duration of sequences in most task categories. The variance between filmed subjects was larger than that between observers in most combinations of parameter and task category. The residual variance, interpreted as mainly being due to within-observer (test-retest) variability, was generally larger than the between-observer variability. This thesis included transfer of research to practice through close cooperation with stakeholders in car disassembly and assembly, their branch organizations and authority stakeholders. The collaboration also included the serial-flow system developer and practitioners through a Reference Group. The transfer of knowledge from this research to the disassembly stakeholders is an underlying principle of the proactive ?intervention?.