Earlier this month, Geoff Parkin shared his first article with WhichPLM. Geoff is an enthusiastic Manager, Industrial Engineer, and Mentor whose specialties include GSD training, and industrial engineering. Having focused on the scope of the Bill of Labour in his last piece, this time Geoff discusses the cost of repair work, and shares his advice for repair training.
It is often argued that “tight” time standards, the pressures of incentive payment and the pursuit of higher levels of efficiency can cause quality levels to drop. I have heard this argument so many times. The fact is that so long as the method set is safe, can achieve the required quality level and an operator can be trained to perform it, then, as long as the time standard issued is accurate for the working method set there is no problem.
On the other hand, with regard to quality and time it is certain that quality errors lead to lost time and thereby reduce earnings and efficiency levels. Quality errors must be repaired and this time for repair is not allowed within the time standard. Repair time is, in my experience, never recorded. As a result nobody is aware of the effect repairs have (and have ever had) on efficiency; worse still, repair downtime adversely impinges on the balance of the line and in so doing reduces the efficiency and earnings of other operators in the same line.
Most factors having an effect on efficiency can be, and in an efficient company indeed are, recorded and reported. But, repair time is not measured. There is a truism, “No measurement, no management”. Repairs will always occur and this critical and adverse activity is not being managed.
What can be done to remedy this situation, can we measure repair times and add them into our management reporting procedures?
I have done some limited work on this in the recent past and found that, on some of the basic small parts and preparation operations, the time to repair can relate to the issued operation standard time. It is likely that on simple ‘unpick and re-sew’ repairs the time taken to repair equates to between 4 and 6 times the standard time issued. The actual repair time is very much determined by the complexity of the operation, the nature of the fabric and how many operations the item has passed through since the fault was sewn.
Measuring repair work with a stopwatch will require a large number of observations and, because of the random nature of faults, would likely take too long to be practical on anything other than the most basic of sewn products. Predicting the time taken by repairs using a pre-determined methods time system would require methods specifications for each type of repair – not really practical given the varying nature of faults.
So what can we do to reduce the impact of repair work on efficiency and earnings?
It is advisable to establish some form of measure for repair time even if this is an estimate; estimates can always be revised as the data develops. The estimated time can be used to track progress as each proposed solution to a problem is applied; a simple multiplication of the number of repairs carried out per operation by the estimate will give an estimated total time lost. Starting in this way will help in developing data gathering and reporting procedures.
The first possible solution may well be to look at operator training. During training we must teach operators how to recognise sewing faults, and a sample book of faults can be created for this purpose. Each page of the fault sample book would contain a swatch with an example of the fault with its description, name, category (critical, non-critical, etc.), likely cause and possible remedy. Using a standard name and fault code for each fault type will ensure that everyone understands what is meant when it is referred to.
Once able to identify faults, operators should also be trained in how to unpick and repair the fault in order to ensure that quality is maintained, needle damage is avoided, and seam strength is not adversely affected.
In addition to recognising their own faults, operators should learn to recognise faults in previous operations within the operation sequence. In this way faults can be picked up before the item has moved too far down the production line. Training records need to be kept separately for this and need to be signed by the operators to acknowledge that repair training has been done and, more importantly, understood.
During the development of operator’s skills and knowledge, line supervisors would need to gain a similar level of ability and gain an understanding of the affect of uncontrolled lost time on their lines efficiency. In the same way quality control inspectors need to understand the impact of repairs on line balance, efficiency and piecework earnings, it is also essential that they should understand the best method of carrying out repairs for each operation. The company must define the best method for inspecting a garment after each stage of manufacture, which will ensure that most faults are picked up on and inspection time is minimised.
Both supervisors and quality controllers will need to learn to work with the operators in a positive and supportive way; a confrontational attitudes towards quality control only leads to resentment. I have witnessed many loud and disruptive discussions between line QCs and operators; these serve no useful purpose whatsoever.
Quality discipline is the responsibility of line management and any such discipline discussions must be carried out off the production line. When presenting faults to an operator the QC should clearly identify the fault and the operator should be asked to agree that the fault is a defined fault, with any dispute arbitrated by the line supervisor.
The next stage of repair training should involve the sewing machine mechanics. It is critical that they understand the nature of faults in order to identify any mechanical causes of faults and possible erroneous fitting/adjustment of folders or attachments. A significant step within the machine repair procedure is for the QC/line supervisor to check the quality of the first few pieces sewn on the repaired machine to ensure that there are no recognisable faults. After this, the repair/refitting of the machine can be “signed off”. Preventive maintenance will reduce repairs caused by machine or attachment failure. Preventive maintenance is essential as machines suffer fatigue, attachments work loose and operators do change settings.
Having established measures to reduce repairs and repair time then it is necessary to deal with data collection and analysis. For this, a standard form must be developed. I propose that the following would be suitable.
Firstly, a style identification header and then, following this, columns headed:
- Bundle identification (bundle number).
- Bundle quantities. This will have come from a bundle ticket or dividers list.
- Operation code, from the BOL (Bill of Labour) or work ticket.
- Operator name, essential for traceability.
- Fault code. Ideally this should tie in with the fault sample book mentioned earlier, and all personnel should have one.
- Number of faults sewn for the above code.
On the face of it, this sounds like a lot of paperwork, but if it cuts down lost time then it is worthwhile.
There may well be an alternative method to the paperwork: a touch screen tablet at each inspection station linked to the factory wi-fi network pre-loaded with the bill of labour, bundle list and list of operators for entry validation would enable Real-Time quality control monitoring; at a number of key stations within factory management could interrogate the system at multiple levels, factory wide, line by line and operator by operator. Within the system setup conditional flags can be set, such as a maximum percentage of faults, in this way the managers terminal would indicate a problem when the condition was exceeded.
At the end of each day a variety of reports can be produced: at a line level a list of operators and percentage faults along with a line total, a league table if you like; at a factory level a list of sections percentage faults along with a factory total again a league table for the factory. As a manager I would expect to see a factory report on a daily basis and be able to “drill down” to operator level to identify problem areas. The same tiered level reporting could be output on a weekly/monthly basis, and quality award schemes could possibly be linked to this as a form of quality incentive.
Additional output from the system could show the most common faults and operations where faults occurred, not forgetting that the estimated time lost due to repairs would tell management the time lost and calculate from this the number of garments lost.
Many companies have invested in IT departments – it should not be beyond these departments to develop such systems in house using database management tools linked to their own existing reporting systems. Imagine the benefits to both the company and its customers from the increased level of control. Imagine if this uncontrolled lost time were under control and reduced, how much extra contribution could be gained?
Setting about a project like this is a big task. If I were asked to take this on, I would want to set up a trial line running the system manually firstly just to quantify the benefits and then to flush out any problems. Given a successful trial then I would want to develop real-time quality control software, which can be such a powerful tool. Real-time QC would allow managers to see trends live and anticipate problems; this can only mean better quality assurance.
It must be borne in mind, the system described is just a tool. This tool will help good managers be better by giving them control over a key issue, it may make average managers become better managers but it will always need someone who wants it to work.