An Interview with Industry Leader Veit GeiseMay 03, 2019
As a former Vice President of Sourcing, what would you say were the main issues in your industry?
I think what made me most nervous on a day-to-day basis, is that ever-looming danger of a quality failure. You open your email in the morning thinking “I just hope nothing failed yesterday.” I could not really control the quality, because there was a lack of transparency in the affiliated processes.
After placing an order, the different production steps happen according to the time and action plan, but despite all our efforts with regards to in-line inspections and spot quality checks, the overall process lacks transparency and is therefore difficult to manage. Hopefully, the Production Manager on the floor, or our QC guys have done their job as good as they can and as good as they normally do it. However, there is an element of the unknown and surprise in there and that made me always very uneasy, following the motto “If you cannot measure it, you cannot manage it.”
So, you came up with an idea to address those issues?
These days, a lot of companies are digitizing the final inspection data. To me, that’s just digitizing the surprise. It doesn’t really help the quality. It’s obviously better to have that data digitized than not, but it doesn’t solve the problem, because it shows me the quality after the fact. It does not enable me to manage the quality ‘in flight’.
What we then did is, we focused on the finishing section of the denim production process. Why the finishing section? Because in denim production you really only know the quality and the problems that you have, once the product comes out of the laundry. It is only then that you see shading, shrinkage effects, broken stitches, raw edges etc. Within the normal production process, the goods then get screened, pressed, monitored, thread trimmed and checked for visual and measurement defects.
The basic idea of The L88P is to digitize that whole process. We decided to organize the finishing lines into an engineered and dictated process flow. We were no longer looking at ’QC by coincidence’, but ‘QC by engineering’. To create the desired transparency, we’ve given all the QC checkers tablets and mounted those tablets on their tables. At the beginning of their shift, the checkers log in with their individual IDs, so we are able to identify and monitor the checker and their resulting work. The individual performance of the QC checkers is largely intransparent in the regular finishing process. The Finishing Manager may have an idea, but the intransparency gets worse as we go higher up in the hierarchy. The buyer/customer has the least visibility of this data or information. The L88P provides this critical performance data and addresses this lack of information.
The most critical question was how we achieve the same level of transparency that we now have created for the workers, for each garment? We decided to sew a very simple QR code into each garment, just below the care label. The code contains the respective PO number and a running ID, such as 1,2,3,4, etc. The checker scans the code with his tablet and the individual tech pack section for the screened item is shown on the screen. The software guides and leads through the process, providing a stable and controlled flow. The checker now checks ‘his’ assigned section of the garment and records his findings in real time in the system. Touch screens and simple color coding make this work easy and error free. At the end of his checking process, the checker saves the work and hands the garment to the next checking station. With this process we now have combined the results of the quality of work of the checking staff with the quality defects rate of the garments, into one set of digitized data.
The real-time available information and transparency from the obtained data makes a big difference. The manager is able to see problems right away and take the respective corrective action. Like in many other companies, we had the rule in VF that our inspector only touches the shipment after it is 100% packed, which actually increases the element of surprise. I only saw the problems after the fact and I totally depended on the quality of work of the checkers and repair workers in the finishing section. I couldn’t manage the process, I only watched the process and was ’firefighting’ the results.
The L88P enables me to see the quality that flows through the finishing unit. I can see a greater accumulation of a particular problem in real-time, trace it back to sewing, cutting, machine setting or operator training and fix it for further cuts or POs. I can see what flows into the box at the end of the packing section. There are no more surprises during the final inspection. I now manage my quality, rather than managing the results of inspections. Knowing what goes into the box makes the final inspection actually obsolete, resolving one of the largest issues in our industry!
What is the name of this product and who is the partner you have found to develop it?
The L88P is the brand name that Pivot88 came up with for the product. We picked Pivot88 at the time, because we were already in the process of digitizing our inspection data with them. They are specialists in bringing software into the manufacturing environment. So, I sat down with Stephane Boivin and told him my idea, and he immediately got the scope of this and got his specialist software people involved. They were able to turn this idea into an amazing product.
How long did it take you to implement it?
It took us about 8 months to design and implement a fully engineered finishing line in a production environment. After about 6 months, the SQL data on The L88P lines showed a 43% improvement over non-L88P lines, and the production efficiency of The L88P lines was 33% higher compared to the traditional lines. Before we started using The L88P, our defect rate in the final inspection stage was about 4.7%. Today, after about 1 year, that SQL data constantly dropped below 2%, resulting in a more than 60% quality improvement!
Now that the implementation process is engineered and tested, the whole implementation will take no more than 4-6 weeks.
What are the benefits of using The L88P?
The main benefit is creating transparency, creating visibility for everyone. What are my defects? What are the issues that are slipping through the line? The machines’ settings may be wrong. If my broken stitches are 80% of my problems that morning, maybe my machines at the bottom hemming operation are not set right, which means I should call the mechanic. There is a lot of dynamic in the finishing unit now where people are taking preventive action and are actively managing the quality, rather than just watching it happen!
The L88P is easy to implement, it’s not a mega investment, and it gives you an instant, massive quality improvement. For example, we had been operating on an AQL of 2.5%, which allows you to ship 5.6% defective garments. Just think about it: You have a 5.6% chance your new iPhone has one or more defects. Or you buy a car and you have a 5.6% chance it has a more or less critical defect. You wouldn’t accept it, right? In our industry, we have been working like that since forever and have come to accept this level of quality.
Before we started using The L88P, our defect rate at the factory level was around 4,7% on average. So, you can see how close to the edge of quality failures (5.6%) we were operating. This led to a final audit rejection of about 10% of the inspected shipments. What happens when you have a rejection of a final inspection in a factory? The factory opens the shipment, re-screens the shipment, represses, repacks and offers for final inspections again. This results in two possible scenarios: They either miss the ship date and air the goods, resulting in a major loss. Or the other scenario is that the factory decides for illegal overtime, because they have already used up all the allowed overtime by scheduling themselves at 8 + 2 hours per day. But running illegal OT to prevent the airfreight losses gets them into trouble during the next CSR audit in the factory. Illegal OT is therefore one of the main failure reasons for factories during CSR audits. So, it is that vicious cycle that we needed to break. And the only way to break it is by knowing what you are doing and being able to manage it.
It seems that using The L88P can save the factory a lot of money?
Absolutely. Because as I said, we used to have a 2.5 AQL, allowing 5.6% defective garments. But we’ve now brought the actual defect level down to below 2%. So, we are going to a quality performance level where failure literally is impossible. This makes me sleep well. There is no more looming danger of failure and air freight and all that kind of stuff. That’s history. Knowing that your quality level is far superior to what the customer expects, enables you to really spend your time on managing to grow the business. That is true QUALITY.
How much does it cost?
It depends a lot on how much you put through, because there is obviously a standard investment cost that comes from the hardware installation. The more units you are running through a L88P equipped line, the faster you’re absorbing that investment. The one thing I can tell you is that it’s a minimal cost compared to the trouble you are saving yourself and your customers.
Do you have any plans for the next step of The L88P’s development?
By digitizing final inspections, you’re‘trying to inspect quality into the garment’. Recognizing this is not the ideal way, we took it the level before, into the finishing unit. But in fact, the garment is already made when it arrives at the finishing unit. So, the next plan of Pivot88 is to extend The L88P into to the sewing section. Capturing the sewing line ID and combining it with the quality data, enables an easy way of tracking defects back to its origin, even if multiple lines are involved in the production of a particular PO.
Today, in a denim production, roughly 15 to 20% defects come out of the laundry, to be found (or not) in the finishing unit. If we bring this down by half to 8% or 10% by better managing the sewing quality, then obviously the finishing can be done even faster and smoother. Pivot88 is now working on this software upgrade.
But in addition to this software upgrade, we also need a qualified engineering staff in the factories. Fortunately, most of the factories now have recognized that they need Industrial Engineers, so you’ll find them in many factories. Not all of them are yet doing real engineering work, but I can see a major improvement in this area over the past few years. It is obvious that you cannot manage factories nowadays without a qualified and well-trained engineering staff.
Does it exist in different languages?
We have it in 9 different languages, so wherever you have your factories, the software can reflect this. And it is very easy to use. We are working a lot with color coding. Red means No, green means Go. We use bold lettering. We want to make sure that it’s easy to understand for everybody, because when it’s easy to understand, it is easy to adopt too.
So, you train the people in the factory?
Correct, Pivot88 will send an implementation team to the factory. But the factory will need to be ready and prepared for their arrival. The space needs to be available. The layout needs to be prepared. The factory will receive a list of the furniture that is needed. The tablets can be leased from Pivot88 or can be purchased by the factory.
Pivot88 will then bring in their engineers to set up, teach and train. It is going to be a 4 to 6 weeks implementation duration and after that, the factory will have their own engineers to run The L88P lines by themselves. The Pivot88 team will always be available to help, but the day-to-day operation will be in the hands of the factory’s own staff.
Any final advice?
From a Sourcing Manager’s point of view, I’d suggest to just take one of your factories (take one where you have a sizeable business) and implement The L88P in one finishing line. It doesn’t cost a ton of money. Do this and let the results speak for themselves. The numbers I shared with you are real numbers, they are facts. And that’s just the beginning of it.
Start to manage your quality improvements, rather than managing bad quality results! It is a lot more fun!
Veit Geise, former VP of Sourcing, VF Corporation