Analysis of Textile Machines Assembly Performance

We were engaged by a German subsidiary in India making textile machinery to analyze the causes of delay in assembly of their machines.

Their normal time of assembly and testing of each machine with about 5000 parts is 15 days while on an average it took more than 30 days to assemble, test and deliver a machine. This problem affected significantly their business performance in two ways:

  • Occupancy of assembly bay, resulting in lower throughput, which could be doubled, if they cleared the assembly bay in normal time
  • Investment in work-in-process inventory by additional 15 days – average of material cost of a machine is about Rs. 2.5 million

The company adopted cellular assembly method with 35 bays of assembly.  About 70% of machine had common parts while 25% varied depending upon configuration selected by customers at the time of ordering. 5% of the parts in some machines were made to design.

The company promised delivery time of six months as they had big order book.

We were engaged to identify sub-assemblies, parts and supplies that caused delay in assembly. The labor availability was excluded from our study as each bay had required number of people with appropriate skill.

We analyzed following data for 3 cycles of assembly:

  • Machine ID, configuration and associated BOM
  • WIP Inventory – used on machine under assembly, allocated and un-allocated on shop floor Warehouse
  • Inventory in main warehouse
  • Purchase order with suppliers
  • Deliveries by suppliers

At the end of the study we provided a report giving details of causes that resulted in delay of assembly of each machine with summary by cause of delay; the key causes were:

  • Sub-assemblies and parts that were not available at the time of assembly; interestingly sum total of value of such parts was less than 3% of the cost of total inventory.
  • Inappropriate allocation of inventories – items were reserved for Machine IDs without considering priority
  • Supply delays by vendors, which were bifurcated into delay in ordering and delay in supply with estimates of average delay days of each part-supplier combination based on history.

The key recommendations were:

  • To assign assembly bays to those machine on order where probability of all parts availability was highest, considering availability and supply in time, based on vendor performance
  • To implement a system of reporting prioritized  list of items, reservations and suppliers every evening so that PPC could focus on chasing the items the next day
  • Review supplier relationships where part-supplier combination delays were higher (or erratic) than specified limit.