On bills of materials structure and optimum product-level smoothing of parts usage in JIT assembly systems

Determining the conditions under which optimum product-level sequences will achieve optimum or near-optimum parts usage smoothing in a just-in-time (JIT) environment has become very important due to computational efficiency issues with industrial scale problems. In particular, this is relevant because while the optimum product-level smoothing problem is tractable, the optimum parts-level smoothing problem is not for large-scale assembly lines. In this article, we build on earlier findings in the literature relating to this issue by considering various bill of materials (BOM) structures - including those used in automotive manufacturing environments. Results from our study indicate that mix of the BOM impacts how well optimum product-level sequences perform relative to optimum parts usage smoothing more than variation in parts requirements. Product composition structure and number of products scheduled are also shown to be very important contributing factors to the above performance. Finally, we find that an increase in the number of products scheduled accentuates the advantage of heuristic part-level sequences over optimum product-level sequences. With the growing use of optional customisation in industries such as automotive, where potentially each customer can order a unique product, these findings are particularly useful in the effective management of inventory.