A socks manufacturing company in India turned to TQM to improve its supply chain. In Part One, the case study describes the process of elminating a significant cost for the company – delivery delays. This week, Part Two of the case study looks at the broadening of the application of process improvement beyond deliveries.

The socks manufacturing company had solved its delivery delays problem and was primed to tackle other issues by expanding its process improvement efforts across the entire company.

As before, the company’s leaders advocated the use of total quality management (TQM) and its seven steps of problem solving combined with just-in-time (JIT) manufacturing principles. The designated improvement team followed this approach:

1. Define the problem
   • Value stream map of current state
   • Desired state emerges from current state
2. Analyze for root causes using current value stream map
3. Generate countermeasures ideas and develop proposed value stream map using JIT principles
4. Test the ideas
5. Implement in production; modify ideas as needed
6. Standardize practice
7. Prepare quality improvement story

Step 1: Define the Problem

Practicing JIT helps remove the “seven deadly sins” of a process:

• Overproduction
• Transportation
• Inventory
• Waiting
• Motion
• Processing
• Defective units

A walk across the company’s production floor showed that work-in-process (WIP) inventory was a problem. The floor was clogged with crates of socks in addition to their warehouse overflowing with other WIP (socks at a separate process stage). No one even knew how much inventory was in stock.

It was clear to the process improvement team members that a drastic reduction in WIP would reduce inventory, release space and reduce transportation. It would also tackle the problems that derived from WIP, such as the unproductive time spent searching for product, stock management, and even reproducing an order when a particular carton or two of a sock variety could not be located.

Step 2: Analyze Why (Find the Root Causes)

The current state value stream map (VSM) is shown in Figure 1.

Adding the inventory lead time between stations (days) and the cycle time for producing one sock (seconds) led the team to the source of the WIP problems:

Lead time = 36 days
Processing time = 24 seconds

Simply stated, a sock that takes a total time of 24 seconds to process waits in inventory for 36 days.

Step 3: Generate Countermeasure Ideas

At this stage, the team applied the JIT principle of demand-pull supply chain management. In a pull-based supply chain, production is based on actual customer orders; no production occurs unless it is in fulfilment of an existing order. Establishing a demand-pull production in the entire sock manufacturing chain meant that the cycle time of each successive stage would be less than the cycle time of the previous stages.

When compared to an upstream machine’s capacity of a 4-second cycle time, three stages stood out as bottlenecks:

To approach the time bottleneck in the linking stage, team members completed a work study of this part of the process.

In the existing process, socks were thrown into a carton in the knitting section for transfer to the linking section. There, the linking machine operator sat in front of his machine and:

1. Picked a bunch of socks
2. Placed them on his thigh
3. Oriented them so that all the toes point to the right
4. Fed them into the machine one by one

The linked sock with the toe attached fell into another carton for transport to the washing station (the socks are washed and dried until slightly damp in an automatic washer in batches of 400).

The time study indicated that the linking machine ran intermittently – for example, it finished 10 socks in 25 seconds while the picking up and arranging of the socks took 15.7 seconds – during which time the linking machine was idle. To improve the cycle time, the following process changes were proposed:

1. The operators in the knitting stage (where the body of sock is made on automatic machines; 10 machines in a bank are operated by one operator) would put all socks in the correct orientation in the carton, eliminating that step entirely for the linking machine operator.
2. An unskilled employee would pick up and hand a bunch of socks to the linking operator.
3. One operator would feed the machine at a rate of 2.5 seconds per sock. One machine would run continuously and result in a cycle time of 3.3 seconds.

The proposed process would use only one machine and two employees to reach a cycle time of 3.3 seconds, compared the earlier process of three operators with three machines and a cycle time of 5.9 seconds. A summary of the time study and the proposed cycle times are shown in Table 2. Note: Tangentially, the goal of reducing the number of operators was also set – from three to two.

Step 4: Test the Ideas – Linking

The proposed ideas were tested successfully as shown in Table 3.

A demand-pull feed from knitting to linking was set up – a feed in which the knitting output moved once an hour to linking. This led to a decrease in WIP from 21 days (see VSM in Figure 1) to a maximum of 2 to 4 hours.

Step 5: Implement in Production – Linking

Implementation of the linking process change was rolled out in stages – first to one shift, then to three shifts and finally to every day.

Washing to Boarding

In the boarding stage, individual socks, still damp and crumpled from the washer, are mounted on heated formers; there they dry and take the “sock” shape. They are then removed for pairing and packing; pairs are prepared, clipped together using a paper label folded over each pair and then packed into cartons for delivery. When needed, pairs may also be packed in polyethylene bags. The loading and unloading from the formers was performed manually.

Prior to the project, six operators were running the cycle for a batch of 30 socks mounted on 30 formers – two mounting, two setting and two removing. The operators were, shockingly, idle for 56 percent of the time, most of which occurred during the 90-second heating time.

A balanced work cycle with only four operators was suggested. As shown in Table 4, this change sought a cycle time per sock of three seconds – but at a minimum needed to be less than four seconds.

Steps 4 and 5: Test the Ideas/Check the Results – Washing to Boarding

The improved cycle worked well. Including inefficiencies in this still largely manual operation, a cycle time of 3.8 seconds was achieved. (The goal of 3 seconds remained, but at 3.8 was still a success at less than 4 seconds.) The improvement is shown in Table 5.

Pairing and Packing

The next stage of the process moved the socks from boarding to pairing and packing. The existing shop floor layout is shown in Figure 2.

A study of the operations in the boarding-pairing-packing area identified a lot of non-value added steps. Table 6 shows the existing process steps, as well as the suggested modifications.

Steps 4 and 5: Test the Ideas/Check the Results – Pairing and Packing

The proposed changes necessitated a change in the layout of the packing tables; they were brought next to the boarding tables as shown in Figure 3.

An updated value stream map is shown in Figure 4.

With the new layout, the boarding operators unloaded pairs from the boarding formers straight onto the packing tables. Once there, the packing operators joined the two socks of a pair with a cardboard folder and pins and then packed them.

Four tasks were eliminated completely from the process: intermediate transport, sorting mix-ups, cleaning fluff and picking up pairs from a pile of socks. The only operations that remained were inspection and packing with a combined cycle time of 10 seconds.

Four operators were deployed to achieve a 2.5-second packing time, which easily improved upon the 4-second cycle time of knitting. Overall, a savings of 11 operators per shift was achieved. The overall benefits are summarized in Table 7.

Step 6: Standardize in Production

All of the above improvements were standardized in regular operations.

Step 7: Prepare Quality Improvement Story

The quality improvement story was prepared and presented to the sock company’s management team. The process improvement team summarized the highlights of its progress as shown in Table 8.

Conclusion

Achieving quality is a never-ending journey – only when an individual climbs one step does the possibility of climbing the next step become visible through the haze of day-to-day problems. The challenge for change agents remains to create an urge for continuous improvement in an organization’s senior management, and then to convert that urge into action. This demands results be proven at every step – fortunately, an attainable task with the methods and tools of process improvement.