A New Approach to Kaizen

A few years Gerard Leone, one of our company’s principals, was asked to sit in on a presentation by architects designing a new building for a hospital client. One of the goals of this project was to make sure that the design was “lean-compliant”. At the end of the presentation, all eyes turned to Gerard, and the CEO asked: “So, what do you think”? And Gerard said, “I don’t like it.”

The issue, as Gerard explained it, was that the Future State Value Stream for this new Surgical Services Department had never been created and used as a baseline guide. While common-sense efficiency had been applied in the proposed design, it was not based on a start-to-finish view of an optimized Value Stream. Gerard went on to explain how this would change and improve the design, and recommended that this multi-million dollar project be redirected to this new approach. The management team agreed, a series of Value Stream activities were conducted in collaboration with the architects, and the end result was a very successful design and project.

This was a “greenfield” project, an entirely new building, which has fewer constraints than a “brownfield” remodeling of an existing facility. There was no Current State Value Stream per se since the building did not exist yet. But what, I thought, if we used this same approach when working on brownfield opportunities, or Kaizen events? What if, in the words of Steven Covey, we “begin with the end in mind”? In other words, instead of starting the Continuous Improvement process with how things are today, we start with how we want the Value Stream to be?

In this article, I’ll be applying this design philosophy to the world of Kaizen, and make the case that this is a better way to apply Continuous Improvement.

What Is Kaizen?

Kaizen is a Japanese word that literally means “Good Thinking”, and which we translate as Continuous Improvement. Kaizen can refer to individual and usually small improvement efforts, and can also refer to larger team-based, multi-day projects. The underlying philosophy is this: unless a process is improving, it is degrading, and so every process needs to have an improvement plan. And it is part of everyone’s job to engage in the effort to make this happen, by suggesting improvements and working on them. Sounds great, right?

Not so fast. Could the very need for Kaizen be a red flag, a warning that something was lacking at the design stage? Here’s what Jim Womack from the Lean Enterprise Institute had to say:

Why aren’t organizations performing lean process design as an integral part of the development process? And are an organization’s skill in after-the-fact kaizen – that is, its talent for process rework – actually reducing the pressure for the hard conversations about lean process development that ought to be taking place during product development instead? Kaizen or Rework? Jim Womack 8/22/2007

Said in another way: can applying Continuous Improvement efforts to a Value Stream or Process actually be a form of waste? Let’s explore this idea more deeply by looking at the traditional way that Kaizen efforts are identified and managed.

The Traditional Approach to Kaizen

The world of Lean is still relatively new, and so nothing is truly traditional yet, but here’s how most organizations go about creating a Continuous Improvement or Kaizen plan. Step one is the creation of a Value Stream Map, documenting the way things are done today, the “Current State”, warts and all. Step two is to analyze the Current State, usually as a team, and identify opportunities for improvement. You might, for example, see that there are portions of the Value Stream that don’t flow well and accumulate inventory. You may also identify processes that have significant quality issues, including scrap or rework. Key bottlenecks in the flow are areas where the team will focus. Based on Lean principles, and common sense, the team will develop concrete ideas for improvement, and add them to the Value Stream Map flow-chart in the form of Kaizen Burst symbols. These prospective improvements, once they are fleshed out and prioritized, will become the basis for an improvement plan. Finally, a Future State Value Stream Map can be created, incorporating the suggested changes into a new flow-chart. This will give you a good idea of the overall changes that you can expect to achieve, and what the new Value Stream would look like.

If you then launch into a series of improvement projects, Kaizen Events and individual projects, then there is little doubt that improvements will be made. In fact, I can’t think of any Kaizen Event that I have attended or led that I would call a failure, i.e. a failure to achieve tangible benefits and improvements. Kaizen Events end with a report-out, a formal presentation of what was accomplished during the event, and the potential benefits are usually impressive. This is a testament to the power of focused effort since the Kaizen team members are relieved of normal job duties when they are working on a Kaizen Event.

The New Kaizen: Process Design

But what if the Current State Value Stream that you are trying to improve was never designed well to begin with? Should you still apply the traditional Kaizen method described above to improve it? Wouldn’t that be what Womack called process rework? Furthermore, if process rework is needed, is starting with the Current State Value Stream Map the best way to begin?

I’m proposing that for many process improvement efforts, what needs to happen first is lean design of the Value Stream. Begin with the Future State Value Stream as the baseline for identifying Kaizen opportunities, instead of starting with the Current State. Here is the difference:

Traditional Kaizen Approach:

Current State VSM -> Identify Improvements -> Future State VSM - make PPT charts

Lean Design Approach:

Future State VSM (Based on Lean Design Principles) -> Current State VSM -> Identify Improvements

Instead of using the Current State as a starting point, you begin by designing the Future State based on Lean Design principles. From there you can analyze the gaps between what you have (Current State) and what you want (Future State), and come up with an improvement plan. The main reason why this approach is preferable is that the Current State of many Value Streams was never formally designed. If that’s the case, then identifying Kaizen improvement opportunities for a sub-optimal Value Stream may be wasting your time, and cementing bad practices in place with Kaizen band-aids. Jim Womack again sheds some light on this dilemma:

As I’ve reflected on this situation, I’ve wondered if the practices of Toyota and other lean pioneers have been misunderstood. Kaizen is an important activity at Toyota and involves all employees. But new processes launched at Toyota are usually extraordinarily lean to begin with and post-launch kaizen is a small part of Toyota’s competitive advantage. Kaizen or Rework? Jim Womack 8/22/2007

Starting with a well-designed Future State does not mean that every improvement project requires a complete overhaul; incremental improvement is still possible and necessary. But it does mean that the improvement projects are based on an optimal state, and not on a pre-existing sub-optimal process.

The Mixed Model Methodology

So what is the Lean Design methodology that I’ve been referring to, as it applies to manufacturing? A bird needs two wings to fly, and the two wings of Lean Design are Line Design and Material Management. The first wing is the domain of Manufacturing Engineering, while the second wing are the Material Flow and Supply Chain disciplines. It should be self-evident that good work flow won’t happen without the necessary materials, and great material delivery is useless without good work flow.

Make no mistake: most manufacturing companies are “mixed modelers”, producing many different but similar products using the same manufacturing resources. The challenges of designing a mixed-model line are substantial, due to the greatly increased level of variability of work content, materials, and processes. A failure to design this type of Value Stream well results in lower output, lower productivity, and increased inventory and working capital. The design skills that your team needs are Mixed Model skills.

Mixed Model Line Design focuses on work flow, and includes Standard Work Definitions, the identification of product families, the concept of Forecast Daily Volume, Takt Time, resource calculations for people and machines, Line Balancing or Workstation Definition, the creation of a Conceptual Layout (and then a CAD-based layout), and the potential use of Simulation Modeling tools.

Mixed Model Material Management focuses on material flow, and includes a Plan For Every Part, the use of pull and Kanban systems, a formal containerization plan, the design of high-frequency delivery routes, the strategic use of Kanban Supermarkets, optimal conveyance methods, and the maintenance of high Inventory Record Accuracy.

Conditions Where This Approach Would Apply

Adapting the approach of designing your Value Stream (or production line) first, and then applying Kaizen to it second, is most applicable when the Current State was not initially designed using Mixed Model Lean principles. And how do you know that it wasn’t? If you have the need for a lot of Kaizen!

If you find yourself in that position, you have two options. We are proposing that you complete the design process in either case, for both work flow and material flow. Your first option, once you have your Future State Mixed Model design in hand, would be to launch a totally new implementation project. This will not be your least expensive option, of course, but it would ensure that all the changes are implemented at the same time, on Day 1 of the new Value Stream. This may be the best option if the changes that you have identified are substantial, and which would be difficult to implement in a phased approach. The number of people, machines, and workstations involved will also make a big difference. The smaller the number of people and machines impacted, the easier it will be to do a total make-over.

The other option is to develop a Kaizen plan and schedule the implementation of your new Lean-based design in a series of events or phases. This approach will take longer, but it may be less disruptive and less impactful on actual production. The goal is to arrive at the same place as the total make-over but in an incremental fashion. Break down the changes that are anticipated into one-week chunks, and schedule them as Kaizen Events.

If you are making these changes incrementally, we recommend starting at the end of the Value Stream and working your way upstream. Think of your new processes as a racehorse, and of the existing processes as a donkey. You don’t want to put the racehorse behind the donkey, so by starting at the end, the faster steed will always be in front.

Getting Started

The Mixed Model methodology, as applied to both work flow and material flow, is a step-by-step approach to Value Stream design that has been tested thousands of times. It’s the way the “big boys” do it, including Toyota, John Deere, and many other Lean leaders. It’s the how-to for designing a Lean Value Stream. But it is something that will need to be learned. Even degreed Manufacturing or Industrial Engineers probably were not trained in the method. So, the first step is to learn it.

Leonardo Group Americas has been conducting Mixed Model Line Design and Mixed Model Material Management workshops for many years, and are currently teaching these courses quarterly in partnership with Toyota Material Handling. We are now launching online versions of these two subjects, hosted on our Learning Management System (LMS) leandesigninstitute.com. You can find out more by clicking on the links below.

Resources:

Info on Mixed Model Line Design Workshop: https://www.leonardogroupamericas.com/go/index.php/mixed-model-line-design

Info on Mixed Model Material Management Workshop: https://www.leonardogroupamericas.com/go/index.php/mixed-model-material-management

Article by Jim Womack on Kaizen or Rework: https://www.lean.org/womack/DisplayObject.cfm?o=753