|
By Brian Dapp
03/31/08
Molecular diagnostics laboratories are faced with challenges including financial
sustainability, staff shortages, space restraints, and relative lack of automation
for highly complex testing. This article will review the principles of lean manufacturing
and how it applies to improving processes in the lab environment. In addition,
it will highlight a recent process improvement project at the Stanford Molecular
Pathology Laboratory, including a discussion on process optimization and suggestions
on how to start your own project.
The Principles of Lean
"Lean manufacturing was first coined in the 1990 book, The Machine That Changed the World, and the term was defined because the authors noted that it uses less of everything compared with mass production in the automobile industry," explains Brian Dapp, Implementation Specialist, OpEx, Inc. (Wichita, KS). "The authors believed that it, lean production, could be applied to any industry and, as result, would have a profound effect around the world."
According to Dapp, there is a lot of confusing terminology that cannot be
defined as "lean." The
following are not lean:
• Six Sigma: "Six Sigma is a tool that is used in lean, but it's a very specific one. A lot of times people will start with Six Sigma and find that it's hard to apply to a process that is highly variable and one for which there is no standard. The first thing you need to do is build that standard into the system and eliminate variation in the form of waste before you can dial in with Six Sigma to a specific problem," he
explains.
• Just '5-S'. "This is a system of cleaning and sorting
through the area, organizing, shadowing, and labeling. It's a tool that's used
by lean, but it should not be confused with the lean methodology."
• A
Kaizen event. This refers to a weeklong project to improve a specific area.
This practice was started after World War II when the Japanese auto industry
was trying to survive. "Again, this is a very specific tool used by lean, but
as far as the bigger picture, simply holding Kaizen events in themselves doesn't
make it lean."
• A project with an end date. "There shouldn't be an end date to a project," he says. "Rather,
there should be a methodology that you manage your business by."
• Easy. "If lean is so great, why don't more people do it?" Dapp asks. "Simply
because lean is not easy."
What is lean?
1. Lean is a strategic business system and a state of mind. "One thing Toyota believes in is using their process as a strategic weapon against their competition. In short, a better process is going to lead to a better product for the customer," he says.
2. Lean is a simple concept that is difficult to implement. "A lot of the principles are common sense, but many of us still aren't doing it."
3. The origin of lean is in manufacturing, but the principles apply to all different businesses.
4. "The single-most important factor in implementing lean is you," Dapp stresses. And, most important, the you extends to everyone in the organization and their involvement.
Conceptually, what does lean mean? "In a traditional hospital layout you have a number of functional departments and if someone wants something from a specific one they have to physically go there…perhaps it's paperwork that needs to be carried down the hall, or a patient who has to move to another waiting area. The goal is to keep everything moving, but all you are doing is going from one waiting room to another. Of course, there are arguments between these departments and budgets to worry about," he illustrates.
"What lean means is that you don't care about these departments. You're going to follow a process, a value stream. In that first analysis, we don't care whose area it went through, what we care about is creating this flow across all the functions regardless of departments. One thing to keep in mind: lean is customer focused and process focused."
Another lean principle is the elimination of waste. "Before you can identify waste, you first need to know what is value added—i.e., what isn't waste.
Value-added activity transforms or shapes material, services, or information to meet customer requirements. Conversely, a non-value added activity requires time and resources but does not add to the customer requirements. "In other words, is the customer willing to pay for the activity? Perhaps it involves the cost of paying someone to run something 200 yards between two departments all the time. If your competition finds a way to eliminate the cost and, as a result, their price is cheaper, the customer is going to go to them," Dapp explains.
Toyota has identified seven classic types of waste by analyzing processes to categorize non-value added activities, which can be remembered by the acronym TIM WOOD:
1. Transportation. "Shipping stuff to different locations. Are you adding value when you're moving specimens between departments?"
2. Inventory. "Keeping stuff on hand when it isn't required. This includes supplies on hand for a project and the supplies you're using."
3. Motion. Excess movement of people or materials within a process.
4. Waiting. Things just don't happen when they should.
5. Overproduction. Doing more than you need to—i.e., the output of a process. "How does it manifest itself in the healthcare environment? Perhaps you store racks of tubes that you never touch just because extras were drawn."
6. Overprocessing. Doing more than you need to within the process. "Converting from a paper system to a computer is one example. It usually means more work because we'll still fill out paper logs and then enter the data into the computer," Dapp relates.
7. Defects. It doesn't meet expectations and you have to do it again.
"And, of course, there's waste of the human mind," he adds. "Many Japanese companies measure themselves by how many productive ideas they get from their employees that improve the process as it relates to the customer. That's what they are always thinking about: How do we improve that process? In the lab, we often spend time thinking about supplies or the next thing that has to get done—not how to improve that process."
To properly frame a lean implementation you have to make each machine and each person a continuous part of the whole process. This requires that you concentrate on the fundamentals, solve the problem from the inside out, and set a foundation for continuous improvement.
The Lean Mission
The lean mission statement is the transformation of the value delivery process—which, as noted earlier, is adding value to the process. The keys to lean include:
• It's devoid of waste.
• Exposes problems real time. "It's hard to stop when there is a problem and truly take action to find the root cause," Dapp notes.
• Deals with the facts. "You need to base decisions on facts and data, not emotions and opinions that often come into play in a discussion," he emphasizes. "If I ask three different people who work in an area about a process, I'll get three different answers about what happens there. There is only one way to know, and that is to go and look."
• Forces problem resolution. "This sounds like a great thing, but it's painful," Dapp admits. "You need to find the root cause to take corrective action. Was it employee error? Or was the error made because the process was set up wrong? This requires a different level of thinking."
• Sustains system-wide continuous improvement. "Always look to get better and never be satisfied," he counsels. "Don't spend your time benchmarking to say I'm just as good or as bad as they are. If there is still waste in the system, there is still room for improvement."
• Encourages legitimate employee involvement. Use your employees for ideas on how to improve the process.
• Meets or exceeds customer expectations. "This is the obvious one—none of us would be in business without the customer."
Lean Training Objective
The key objective is to develop the ability to recognize and eliminate waste. "You need to have the courage to call it waste, the desire to eliminate it, and then act to eliminate waste," Dapp says. "Understand that waste raises cost, produces no corresponding benefit, and, in the long term threatens all of our jobs. Sometimes we get shortsighted. My argument is that if the competition finds a way to do it better and charges less, eventually what's going to happen? One of my favorite sayings from going through lean training is, you get what you expect, and you deserve what you tolerate. If you're happy with how it is today, that's how it's going to be. If you expect something better, then you're going in the right direction."
What's possible when you apply lean in the laboratory? You can reduce turnaround times by 70 percent, increase productivity improvements by 35 to 50 percent, save 30 percent on space, reduce inventory and supplies by half, standardize work practices, reduce errors or the potential for errors, and implement performance measures.
"One of the lean concerns that always comes up is that making improvements will result in fewer jobs. That's the first thing people usually start thinking about, but that's not what lean is about," Dapp points out. "Lean is about improving the process and eliminating waste—it's not about eliminating people. If we can free up people, they can take on additional work and grow the business. That is one of the issues that has to be dealt with up front and explained to people. Henry Ford, for instance, knew this to be true."
Another frequently asked question is when is lean going away? "The answer is that it doesn't, it's a never-ending journey," Dapp notes. "Lean is a different way of thinking and running a business. You can't go through lean one time and be successful in the long term."
A Brief History of Lean
"While many people are familiar with Henry Ford, the current principles of lean can be attributed to Taiichi Ohno, an engineer at Toyota. In August 1945, Kiichiro Toyoda, president of the Toyota Motor Company told the company that they needed to catch up with the Americans within three years, otherwise the automobile industry in Japan would not survive. Ohno visited Ford and, while he was impressed, also believed that Japan needed variety in their products. People used to say about Ford, you can have it any color…as long as it's black. Ford figured out how to do one thing very well, but that wasn't what Toyota needed. They needed trucks, vans…different types of product," Dapp relates.
"While in America, Ohno also visited supermarkets. In the 1950s, people in Japan did their shopping at mom-and-pop stores—you got bread at one shop, meat at another, and so on. When Ohno walked into an American supermarket, he saw one place where he could get whatever he wanted and items were continuously restocked. This gave him the idea of a manufacturing system that relied on somebody wanting something specific, getting it, then making another. Rather than have a whole bunch of black cars coming down the line, the customer could choose. This was the birth of just-in-time manufacturing. If he had just benchmarked to be as good as Ford, then Ohno would not have been thinking outside the box and developed a different approach to the process. As Albert Einstein said, The problems that exist in the world cannot be solved by the same level of thinking that created them."
Applying Lean
"There is no one method or procedure that could be applied in every environment, at all times, or under all conditions to create lean," Dapp notes. "However, there is a basic pattern of approach, which includes a clear vision of the future; elimination of waste; clearly defined goals and objectives; excellent project planning and coordination; excellent command and control of the fundamentals; systematic and methodical approach to establish the facts; constant and unyielding pursuit of excellence, and step-by-step development and constant implementation. Finally, lean has to be directed by a strong leader who can get everyone on board."
According to Dapp, making the changes required for lean requires a working knowledge of lean principles and techniques; rapid systematic analysis and implementation; ongoing improvement; focus; and detailed, hard work. "Most important, you need to solve problem from the inside out. This means that there's no canned answer as to what's best for an organization striving toward lean principles. There is no machine, piece of software, or one method that is going to get you where you want to get to."
The 12 Operating Principles of Lean
The 12 operating principles of lean can be applied in any organization.
1. Standard work methods. "This is not the same as standard operating procedures," Dapp notes. "The dictionary defines standard as an acknowledged measure of comparison and criteria; widely known and accepted; and used as the basis for a system of measures; and a degree or level of requirement, excellence, or attainment. We agree that we want people to achieve this standard. In reality, however, people have different ways of doing things while they may come up with the same result. It's not going to be easy to convince someone who has been doing something the same way for 20 years to change their habit. You need to have a standard that you can hold people accountable to."
The essential elements of standard work include:
» Process Sequence: This includes how long it takes, the turnaround
time, and when it should occur.
» Workstation Layout. "You need to have the workstation defined the same for a specific task."
» Quality and Safety of the Process. "You need to define each and build it in to your standard."
» Operator Walk Patterns. "If you have everything where it is supposed to be and you have a process design, you should know where they have to walk to get between points," he notes.
» Standard Work in Process. If true one-piece flow can't be achieved, then you have to design a process with enough work in process to meet customer demand; and it should be defined in your standard work. "As an example, we know that if an instrument uses high-dollar controls you're going to run 100 samples at a time. Or, you're going to wait until you have four hours worth of work and then run it. These are definitions of standard work in process," Dapp illustrates.
» Machine Cycle Times.
"Standardization promotes quality through simplicity and consistency," Dapp says. "However, this principle is so obvious that sometimes it gets overlooked. The process is so well defined and consistent, and the supplies are always there when you need them. You don't have to go search or try to figure out another way of doing something. Eliminate this variation and you have a much higher quality standard."
2. One-piece flow. The following flow chart illustrates the results of running 10 specimen samples
using batching (traditional method) compared to one-piece flow.
"Let's assume we have three technicians each performing a different process. In the laboratory this could mean phlebotomy, accessioning, and analysis. Also assume that each process takes 1 minute, there is no move time between operations, and you're processing 10 specimens before handing them off to the next process. So a phlebotomist goes to the floor and draws 10 specimens before handing them to accessioning, which then accessions and hands them to the lab. The lab performs the test and provides results. From the time of the first draw until the first result it will take 21 minutes—and we don't even know if this was the first specimen or last specimen drawn. Finally, after 30 minutes from first draw you will have results for all specimens. This method is filled with potential for errors as labels and specimens can get mixed. There is also a great deal of variation in turnaround times potentially ranging from 12 to 30 minutes," Dapp explains.
"In a one-piece flow environment, once the first sample is drawn, it is immediately handed to accessioning, which performs their work and then immediately hands it to the lab. In this example, the first result is achieved 3 minutes from draw through result and the last result is entered after 12 minutes. There is no chance to mix labels/specimens and the turnaround time is the same for each. This may seem unrealistic right now, but it should be what you are trying to achieve," he notes.
According to Dapp, to achieve one-piece flow you have to address many of the reasons for batching, including setup time, location of equipment, expensive equipment, isolated work islands, lack of cross training, and it just "feels" more productive.
3. Standard work in process. "Basically this identifies the minimum amount of work in process that is needed to maintain your standard work in the way you've set up the system," Dapp explains. "If, for example, can't run one-piece flow, you might wait for the fourth sample to hit and then load the centrifuge. It's knowing where to make that determination to batch; it can be in the volume of that product or in the time frame. Some things are critical, such as cryogenic preservation of stem cells. Once you add dimethyl sulfoxide (DMSO) you have two hours to start the freeze-down process. So your standard amount of work in process becomes two hours. Once the first sample goes on the DMSO, you keep running samples until you hit two hours. Then, no matter how many samples you have, you have to start the cycle to freeze them down. It can take those two different forms—the number of samples or how long it takes, but the number is known and you can design and build it into the process."
4. Takt time (demand) pace. This can be most easily defined by the following equation:
Takt Time = Allocated Time
Required Quantity
Takt time is the amount of time or speed you need to produce the required results that is driven by the customer. Takt comes from a German word for the beat of an orchestra conductor's baton, which sets the pace. "If we know we have standard work in place, we know how long it takes to do a task. We know our demand data, so we know how much of it to do. The frequency of the maximum demand can also be used to calculate the peak of the system. It's really simple math and it will also tell us how many people you need to do the job, known as staff to demand," Dapp says. "Then you can look at the way you're set up to determine whether your staffing requirements fit the demand profile—i.e., too few or too many people."
5. Balanced distribution of work. Equal division of labor within a work group. "For example, if there is 100 minutes of work, and there are four operators, each person should have 25 minutes of work This is a fair system and failure to abide by it can breed contempt," he notes.
Dapp adds that Parkinson's Law says that work expands to fill the time available for its completion. In other words, you can assign the same task and allocate a different amount of time to two different people to complete it, and each person will complete it based on your expectation—whether it's two hours or eight hours. In a lean layout where all the work is distributed evenly and you apply Takt, if one person takes either more or less time than has been allocated to complete a task, that might point to inefficiencies in the process.
6. Visual management control. You need to clearly and visually notify management if the work area is normal or abnormal. "This can take the form of shadowing and labeling, andon signals (a light that is used to notify management, maintenance, and other workers of a quality or process problem), downtime clocks, or process control centers. Downtime clocks are helpful to determine how much time is being lost to a specific problem. The data can then be used to determine where to focus additional improvements. Process control centers provide a summary (via a blackboard or plasma screen) that displays metrics for staffing, production, etc. Shadowing and labeling helps identify where items belong—from a timer to a trashcan.
Visual management control follows the 5S methodology:
Sort (seiri): Clearly distinguish what is needed and what is not. Remove
what does not support the Least Waste Way.
Set in order (seiton): Organize the way things are kept, making it
easier for anyone to find and return items to their proper location in the
sequence used.
Shine (seiretsu): Keep things clean—floors, machines, desks, files,
equipment—neat, and tidy.
Standardize (seiketsu): Maintain and improve the first three S's (what
causes deterioration).
Sustain (shitsuke): Achieve the discipline and habit of properly maintaining
the correct procedures.
7. Multi-process handling. "The general rule
in labs is one instrument, one person; person watches the machine run. However,
this is not efficient. You need to separate the person from the machine so
he or she can do multiple things."
8. First-in, first-out materials. "Always use the oldest first—first-in, first-out system at point of use (POU), which will eliminate throwing out expired material," Dapp says.
9. Pull material replenishment systems. "FIFO is only the first half of the equation," Dapp explains. "The second half is to establish a pull material replacement system that ensures that the two rules of material control are always maintained—don't run out and don't carry more on hand than you need. Your system can be as simple as a card at the bottom of a supply bin that lets you know that replacements are necessary."
In fact, Dapp notes that a lot of time is wasted on supplies. A supply sits in inventory until it can be used and then billed. All other supplies that are unused represent a waste of money—which is inventory, one of the seven wastes. "The faster you can turn supplies around, the better—and you can accomplish this by getting materials to the point of use for technicians so they don't have to waste time doing it," Dapp says.
10. Rapid changeover. "It's important to try to reduce the amount of time it takes to set up a machine—and that's one of the reasons for batching. A lot of times people think they don't have enough centrifuges. However, how often is that centrifuge actually cycling; how often is a person actually loading and unloading—doing other things while the machine isn't running? One of the lean principles is to figure out how to keep the machine running efficiently and do other activities while it's running. This is what set-up reduction is all about and it can buy you a lot of machine capacity," Dapp notes.
11. Mistake proofing. This is defined as eliminating the chance of defects by providing immediate feedback that an error is about to occur—i.e., exposing problems in real-time and eliminating the production of waste. "Mistake proofing is not only about preventing the error from occurring, but changing your mindset to look at the process so it can be eliminated completely," he points out.
There are several approaches to quality control, which are summarized from most to least effective as:
• Physical
barrier that prevents an error from occurring
• Visual controls with objective
readings
• Subjective visual comparisons
• Training by conducting periodic honest reviews
• Reminders posted at the point of use
• Standard operating procedures
12. Right-size equipment: "Nine times out of 10, people buy big, expensive machines that are capable of doing everything. In a lean environment, we prefer to look at equipment in terms of what we would need ourselves to do the job. For example, you're looking to invest in a centrally plumbed aspiration system throughout a new laboratory. This is a very expensive investment that will require maintenance. If the system goes down, the whole lab goes down. In right-sized equipment, you can purchase a smaller version of this system that plugs in at the point of use. You can move it, rearrange it, and create new layouts pretty easily. And it doesn't cost near as much as a centrally plumbed system."
Success in Lean
"Success in lean is the complete command and control of these operating principles," Dapp concludes. "Failure is dumbing down the defintion of success until success and failure are synonymous."
|
