Lean manufacturing in custom metal fabrication: How to define the current state

You are like many custom metal fabricators. Challenges abound. Problems nip at your heels all day long. But you have made the decision to take the first steps toward significant improvement. The commitment to begin the lean journey is in place with the leadership team.

Your company has lots of issues, but cost and lead-time problems seem to be front-and-center factors that get in the way of your company’s good performance. After careful consideration, you decide to do a lean pilot project focused on the frame assembly in the welding area to improve performance and gain traction for the lean journey.

But what do you really know about the current state of the frame welding and assembly area? You’ve heard opinions and anecdotes, and you’ve also heard “this is the way we’ve always done it” too many times to count. So how do you make sure you really know?

Step 1: Go See

To start, you simply go to the frame welding and assembly area and look. Really look. Even though you walk through the welding department every day, you rarely take the time to observe.

As you walk into the shop, you recall a conversation with the production supervisor responsible for the low-volume, high-product-mix press brake operation. The supervisor said he recently observed the changeovers and saw one that required the operator to “loosen, tighten, loosen, and tighten” approximately 30 set screws with a T-handle Allen wrench. This contributed significantly to the 45-minute changeover time. The supervisor said he had supervised (that is, walked by) this operation 20 times a day for three years … and had no clue about how labor-intensive changeover really was. He simply had not taken time to “go see.”

As you walk to the welding area to “go see,” you wonder what operational deficiencies would become apparent. Were the deficiencies there yesterday, last week, and last month? Probably so.

When you reach welding, you start to see problems immediately. The flow is erratic. The starts and stops are visible. One station is buried with work while the next is starved. You watch one skid of material enter the work area and follow it all the way through the process.

The flow has little logic to it. People put material wherever there’s room, and the next person who needs the material has no idea where to find it. There is way too much unnecessary movement, and the erratic flow pushes piles of work-in-process (WIP) ever higher.

The engineer and production manager arrive, and after all three of you stop long enough to “go see,” it becomes obvious: WIP levels are being driven by an imbalance between operations. The operation with the faster cycle times is burying the next operation, and any open space is fair game to store the WIP. You look at your colleagues and wonder why nobody saw this before … or if they did, why wasn’t something done!

As you observe a changeover from one product to another, you see the welder spending approximately 15 minutes rooting through a disorganized collection of fixtures. All the fixtures, unlabeled and unidentified, are stacked and stored haphazardly, and the welder is thoroughly frustrated by the time he finds what he’s looking for.

Something else catches you completely by surprise. Between the three welders on first shift and the one welder on second shift, you discover that four different welding parameters are being used to produce the same product, even though the product has a specified weld procedure. No wonder you have so much variation in product quality.

Another observation: Each welder performs their work in different sequences. Two of the welders start in the middle, while the others start at the ends of the assembly. The production manager points out how this could result in variation caused by different patterns of heat distortion and warping. You begin to understand how loose your company’s standards are!

After a full two hours standing in the area and observing, it becomes apparent that the “go see” exercise is essential to developing an objective understanding of the current state. In fact, you soon realize that the entire leadership team needs to learn to “go see.”

Make Sense of the Data

The next step in defining the current state requires operational data. You are very familiar with the plant’s outcome metrics. How many units did we ship? Was the plant profitable last month? What are customers complaining about now? But you recognize that all this tells you little about the condition of the process in the welding area. So what operational data is available?

The quality data provides insight into the final product shipped to customers, but it doesn’t tell you about the quality of the weldments leaving the welding area, at least not directly. The best information you find is the scrap count. Based on what you saw during the “go see” exercise, you know welders spend time doing rework, but the rework time is not accounted for. So you assign the production manager to find the best estimates shop personnel can provide.

Shop travelers and prints accompany the parts as they wind their way through the process. Direct labor time is reported against the shop orders, so data is available on when jobs were finished, how long they took, and what quantities were produced. Each operation also has a planned start date, so you can determine if a job started on time, early, or late. In the welding area, it seems that most jobs start late. Does this mean the jobs are late before they even start? To find the answer, you need more information.

Customer feedback comes in several forms. Some comes from an annual survey your company does with its customer base. It’s not very helpful for this pilot project, because the data is too broad. But could you obtain anecdotal input from customer service personnel? They have the most direct contact with customers, after all. Can they provide information they already have or make a call to gather input directly from customers? You’ll need to follow up on this.

Basic operational measures tend to be focused on time, quantity, distance, and floor space. People working closest to the floor can relate to these measures much better than to financial measures like monthly profit/loss and return on investment.

So you send the engineer and production manager back to the welding area to gather two operational measures. One is to step off the travel distance from the end of the operation before welding (the press brake) to the beginning of the operation after welding (paint prep), plus all the movement within the welding area. They’ll also draw a spaghetti diagram to illustrate the flow.

You also need to know how many times a part is handled. During the “go see” exercise, you noticed that operators are spending a lot of time picking up and putting down parts. It seems like a lot of unnecessary handling and movement, but that’s just an assumption. You need data.

These operational measures will help you really understand the current state. You could pursue others, but these will give you a good start. You also are beginning to understand how a few effective measures can help take a lot of the emotion and judgment out of conversations.

Conclusions About Current State

You and your team look at all the findings from the current-state analysis in the welding area. You begin to see why the costs incurred in the welding process seem high. Actual costs are running approximately 25 percent above expected costs (that is, standard costs). Although the overall lead time is about four weeks, one full week of that total time is spent in welding.

So now you know why the cost and lead-time metrics seem so out of whack. The welding area’s flow is a mess, tooling is disorganized, there is way too much WIP, and there are constant disruptions from hot jobs.

You now have the information you need to give to the project team that will lead the improvement project. You could decide to conduct an informal improvement project that may span several weeks with part-time effort from the team members. Or you could conduct a formal five-day kaizen event to attack the welding area with overwhelming force! Either way, you know that the project is justified. So let’s get on with the improvement!

Jeff Sipes is principal of Back2Basics LLC, 317-439-7960, www.back2basics-lean.com. If you have improvement ideas you’d like to read about, contact him at jwsipes@back2basics-lean.com or Senior Editor Tim Heston at timh@thefabricator.com.