Cellular manufacturing systems considering human characteristics

三十多年来，成组技术和单元化制造一直被认为是解决许多制造问题的方法。实施单元制造通过减少安装、物料处理、在制品和制造提前期，帮助许多组织迈向精益制造。这些好处通常是通过将具有相似特性的部件和产品分组到族中，以及将共享零件族工艺的不同机器分组到机床单元中来实现的。一个可能被忽视的问题是蜂窝制造的人的方面。很少有研究通过对这个问题采取顺序的方法来解决将操作员分配给单元的问题。也就是说，形成了第一个零件族和机床单元。然后，将运算符分配给单元格。然而，这种方法可能不能完全解决问题，因为许多操作员可能被分配到不适当的牢房，他们的技能没有得到充分利用。操作员可能有不同的技能和对工作的不同偏好。此外，他们的生产率取决于他们的工作经验。如果不考虑这些重要因素，将导致操作员失去动力和兴趣，并降低系统的生产率。本研究提出了一种在设计机床单元和零件族时同时考虑操作员的技能匹配以及他们的经验和偏好的方法。拟议的研究将审查现有的绩效衡量标准，以便将其扩展到感兴趣的多维问题。它还将回顾许多现有的机床单元/零件族形成技术，以研究它们在机器-零件-操作员分组问题中的性能，并将提出具有广泛适用性的广义分组问题。使用建议的绩效模型，将选择最佳分组方法。

For more than three decades, group technology and cellular manufacturing have been considered as solutions to many manufacturing problems. Implementing cellular manufacturing has helped many organizations move toward lean manufacturing by reducing setups, material handling, work-in-process, and manufacturing lead times. These benefits are usually achieved by grouping parts and products with similar characteristics into families, and grouping dissimilar machines that share the process of part families into machine cells. One issue that is likely to be overlooked is the human aspects of the cellular manufacturing. Few studies have addressed the allocation of operators to cells by taking a sequential approach to this problem. That is, first part families and machine cells are formed. Then, operators are assigned to cells. This approach, however, may not fully address the problem, as many operators may be allocated to inappropriate cells where their skills are not fully utilized. Operators may have different skills and different preferences towards jobs. Also, their productivity depends on their experience on the job. Failure to consider these important elements will result in loss of motivation and interest for operators, and low productivity for the system. This research proposes a simultaneous approach that considers operators' skill match and their experience and preference while designing machine cells and part families. The proposed research will examine the existing performance measures in order to extend them to the multi-dimensional problem of interest. It will also revisit many existing machine cell/part family formation techniques to investigate their performance for the machine-part-operator grouping problem and will propose a generalized grouping problem with a wide range of applicability. Using the proposed performance models, the best grouping method will be selected.