Systematic forecast of human reliability in manual assembly operations in serial production

• Applicant: Professor Dr.-Ing. Robert Refflinghaus
• Subject Area: Production Systems, Operations Management, Quality Management and Factory Planning ; Human Factors, Ergonomics, Human-Machine Systems
• Term: from 2014 to 2018
• Project identifier: Deutsche Forschungsgemeinschaft (DFG) - Project number 251787698

Assembly plays a leading role in the German economy, especially in terms of added value. Depending on product characteristics assembly operations take up to 70 % of the total production time. In recent years, shorter product life cycles and increasing quality requirements of customers led to increasing demands also for the planning process of manual assembly operations.A major goal of quality planning deals with creating strong production processes and reducing the number of errors in production. Transferred to the field of manual assembly processes, this means, besides mastering the technical complexity, also to reduce the human error probability of the employee. The early detection of quality-critical operations therefore is a key element for generating stable and efficient production processes.The evaluation of weaknesses of the production process is usually limited to technical and economical fields. Currently, the planning process of manual assembly processes is primarily focused on an optimization of time scheduling. For this purpose, assembly planners generally use systems of predetermined times for calculating the time that is required to perform a manual assembly operation. Despite of first scientific approaches developed in the previous project MTQM a systematic consideration of human reliability in assembly operations so far is neglected in the operational practice.In this research project, a method will be developed that has the potential to predict human error probabilities in manual assembly operations systematically. The starting point of the development process are the results of the previous project MTQM and also process analyses created by using systems of predetermined times. In addition to the calculation of predetermined times, these process analyses are used to assess the complexity and thus also the error rate of individual sequences of action of manual assembly operations by determining cue motion values for individual task items. With the gained knowledge a systematic procedure for evaluating human reliability in manual assembly operations will be developed. The above mentioned procedure will be based on an Expert System For Task Taxonomy, which for this reason is especially adapted to the certain characteristics of the considered application field. As the human reliability of manual assembly operations is strongly affected by learning curve effects, the influence of experience of the employees and its impact on the amount of resulting error probabilities will also be examined while developing the new method. Due to the prospective determination of human reliability in manual assembly operations the assembly planner will be able to implement necessary adjustments already before start of production. With this procedure costly redesigns during ongoing operations can be avoided.

• DFG Programme: Research Grants