Gemba Panta Rei

Iain Johnstone
Operations Consultant

TPS is a set of tools and philosophies that model the perfect world. What some organizations, their management team and Lean consultants forget is that we need to strive for our ideal world. The perfect world is out there and whenever I am training anyone on TPS I teach them about the perfect world however, I always tell them that they will never make it to the perfect world. It is a model that we need to understand and strive for but we should understand it is unattainable. If anyone ever tells me they are operating in the true model of TPS and have made it to that perfect world I would have to laugh.

In talking to a VP of Manufacturing who had attempted to implement lean with the help of a consultant in the past, they agreed with this concept. The company, a furniture manufacturer, had short lead-time (5 days or less between customer request and shipping), high mix (400 different products and components) and a very unstable volume (one day they would have 6 trailers of deliveries, the next they would have less than a quarter of a 45 foot trailer). Each machine that the component went through would have a set-up time in the region of 30 minutes with a cycle time of 10 to 30 seconds and each component may go through 5 or 6 machines with around15 components per assembly – not your ideal scenario for single piece pull at takt time.

The consultant had trained both the company’s leadership and the operators on the floor. He had explained the philosophies of Takt, Flow and Pull and was driving the manufacturing company towards a JIT environment. It would appear that his drive was towards the perfect world with little thought given to the consequences on the rest of the organization or how they would make this leap. Previously the environment was batch & queue, running orders of 300 parts through a machine in order to minimize the set-up time cost per part and running one or two different parts on that machine each day. The consultant pushed the company towards batches of 25 to allow them to be able to process at a higher velocity and increase the number of different parts made per day. Great you might think….we are aimed at the principles of JIT and utilizing some of the principle of Heijunka. Once they had been able to bring the batch size down to 25 the plan was to reduce it further and further and further until they were at single piece flow. Unfortunately this blinkered approach caused costs to ramp up significantly and very soon the VP of Manufacturing had to say stop, this is not helping our business it is hurting it!! It sounded like the consultant knew the direction he should be taking the company but hadn’t considered the best way to get there, perhaps some SMED would have been the better first step?

The consultant took a very rigid position and told the Executive VP and President of the company that the VP of Manufacturing was a dinosaur and had to go because he didn’t understand the principles of lean & JIT and was so ingrained in the batch and queue system that the company would never change while he was in place. Fortunately the senior management took the side of the VP of Mfg and the consultant was the one to go. Even today though, the senior management will tell you that they tried lean but weren’t able to make it work. I saw many places where some of the lean principles (Jidoka, pokayoke, standard work, visual management etc) had been applied to the facility.

What both the consultant and the senior management have missed is that link between ideal and perfect worlds. We should understand the perfect world and hold it up as a light to guide us but we need to temper that with a modicum of reality to find our ideal world.

Posted by Jon Miller at 12:45 PM | Permalink | Comments (0)

Iain Johnstone
Operations Consultant

During a recent trip to a client, myself and a team of 5 from various manufacturing departments began mapping the first three weeks of their build process. The product is very complex and labor intensive with an average of 5500 hours of cycle time during an 18 week lead time for a “standard” product. Due to the complexity of the product and the number of people involved (250 direct employees), I had previously deemed a fishbone diagram the best way to map the processes and their interactions & dependencies, this worked very well for the first two weeks of the process that had already been mapped. One of the reasons for mapping was to help us identify which processes should be the subject of kaizen events in the future.

Initially the team had some concerns about being able to accurately show how a coach is built because of the complexity and the number of people involved in each of the tasks but after learning some lean basics including TFP they felt a little more confident about being able to see what is going on. The power of mapping soon shined through once the team began to get tasks down on paper and realized there were so many things they didn't understand about what each department and operator does and immediately were changing the order of tasks to maximize the amount of work completed as sub-assemblies away from the coach whilst making sure there was always as many people working in the coach as possible. As the map took shape more and more people were brought in from the floor to add their expertise in particular tasks and again they were pleasantly surprised by the amount of information we were able to get on paper and how clear and useable the map was.

Once the map was complete I converted it into a Gantt chart for the manager to use. Previously the only “plan” in place for each department was a list of tasks that they understood needed to be completed in the next three weeks, there were no clearly defined responsibilities as the process had evolved over time and people just knew what needed to be done next. On the Gantt chart we identified each task, who was to do it, what time they should start and how long it should take. From this we were able to link tasks that were related or dependant on one another together and everyone could see how they would impact the rest of the department if they did not complete their tasks to the plan.

Initially there was a great deal of concern from some people that they were now being locked into the schedule and no longer had the flexibility that they had enjoyed up to this point. My response was that “you cannot manage what you cannot measure and without a plan to work with, how could you ever measure if you were on track!” Although most understood this principle there was still some reservations however, once they started using the Gantt chart most people felt the benefits pretty quickly. Not only were they held to the plan but so was everyone else so the tasks that they had to wait to be completed in the past were now getting completed on time and in the right order. They had also leveled their tasks out over the entire three weeks to make sure they didn’t have a big rush at the end to get everything finished. Creating the fishbone map had allowed us to rearrange the tasks and get three weeks worth of work completed in two weeks by eliminating dead time and reducing lead-time. Ultimately within a month of using the Gantt chart the team had been able to take an additional two days out the their lead-time by fine tuning the process through use of a kaizen newspaper.

It just goes to show, we don't know what we don't know until we map out what is happening and can see the big picture.

Posted by Jon Miller at 12:43 PM | Permalink | Comments (0)

It's interesting how things come in threes. Recently the issue of stability and how it affects successful Lean implementation came up three times in rapid succession.

Following up with a client of ours who is a candy factory (very seasonal), we learned that they were plagued by material outages, unplanned staffing shortages, shipping errors, and over 130 minor stops on their main line during peak season this winter. Part of this was that their efforts to go Lean had "exposed the rocks" by lowering inventory and excess staff, and part of it was that they were now measuring these losses and can put a finger on what is preventing flow and on-time delivery.

Now that these sources of instability (losses in manpower, machine uptime, and materials) have become clear, they can address these and continue to build their Lean enterprise.

I spoke with another one of our clients in the automotive industry is facing the challenge of delivering products to customers in the face of instability. Their orders are 60% higher than even their most optimistic forecast. Their main assembly plant is in a low unemployment county and it is demanding, physical work. This results in their absenteeism and turnover accounting for 10% to 15% manpower shortage on any given day. They have long lead-time raw materials and schedule changes that happen daily due to a lack of methods and standards in customer service.

While there are Lean tools (OEE, Takt & cycle time balancing, Kanban, Standard Work) to remedy these situations, once you are caught with a backlog and one or more of the elements of manpower, machinery, methods, or materials missing, it can be very hard to find the time to do what's needed to catch up.

In sharing our experiences about applying Lean in hospitals with Cindy Jimmerson of Lean Healthcare West, she tells me that "No hospital unit is stable". Cindy brings 25 years of experience as a nurse and has important perspective on what it takes to make Lean work in a hospital. She means that turnover of staff, spikes in patient census (what we would call an inability to level the load), and a lack of standards in among physicians in methods of patient care the also in information flow processes supporting patient care requires a Lean implementation approach that is tailored to healthcare.

I learned a lot from a VP at a client of ours named Paul. Years ago Paul taught me an important lesson. Our client's management team was struggling to embrace the changes we were introducing via Lean, because their behaviors and habits had not changed. While they saw the benefits of flow and pull, when absenteeism spiked up, or there was a supplier delivery problem, they went back to what they knew, traditional batch & queue.

Paul said "Jon, you have to practice. You can't just play." We all want to scrimmage and play ball rather than practice the drills over and over again when we're on the field. Yet if we don't have the fundamentals of throwing, running, catching, and blocking, we'll lose on game day against a team that has been practicing all week rather than just playing. Paul was right, we were "playing" by focusing on implementing the Lean tools, rather than creating a foundation of stability first. Practice is less glamorous, and less fun perhaps, but vitally important.

A simple assessment of stability for a manager or a Lean consultant is 滴ow often do I have enough resources (manpower, materials, equipment availability, and stable and repeatable processes) needed to fulfill the request of my customer?" Most organizations measure on-time delivery, absenteeism, uptime (or OEE), and process capability. These are all good measures to answer that question.

Once you identify what "enough resources" means to fulfill customer requests (volume, mix, promised lead-times) then you will be able to manage and improve these metrics through Gemba Kaizen. Ten more laps before the game.

Posted by Jon Miller at 10:17 PM | Permalink | Comments (0)