Journey to Lean: Making Operational Change Stick

Journey to Lean: Making Operational Change Stick

---

Managers who want their employees to engage in the change effort must first be willing to engage with them. What surfaces during the dialogue should not be dismissed as carping, but should be translated into positive action, with employees encouraged to help devise solutions to the issues they raise. There are No Shortcuts Most lean programmes include three ingredients: Although all three can make valuable contributions, they need to form part of an integrated transformation programme to succeed.

Wrongly applied, they can cause damage and lead to disappointment and cynicism. Taken in isolation, each has its drawbacks. Workshops teach tools and techniques rather than help create a company-wide system for running a business. In benchmarking visits, observers tend to see only the physical aspects of lean operations; the underlying principles remain hidden and unexplained. A typical kaizen workshop would focus on just one of these steps, and seek to make a dramatic change in it through a focused effort over a week or two.

A multi-disciplinary team would be assembled specifically for the workshop. By drawing together a handful of people from various functions for a short period of intense activity, such workshops create a temporary organisational form focused on a single narrow task. When the workshop members go back to their everyday jobs, the process operators — the people actually responsible for implementation — lose the support they need.

None of the business systems have really changed, and the process quickly reverts to its former state. Benchmark Visits In our experience, it can be very difficult to make sense of what you see on a benchmark visit. Take a tool such as kanban. For more on kanban, see Chapter 3. You may know that the kanban is designed to trigger upstream production and delivery, but what you see is simply a small card attached to a box of parts.

Moreover, a visit is just a snapshot, not a moving picture. Lean practices are not things that can be copied from one place and pasted into another. A lean system is more like a living body than an inanimate object. It links all the various parts of an operating system into a coherent whole. This is why isolated lean initiatives so often fail. Recruitment from Lean Companies Employees with hands-on experience of lean operations tend to come from production environments. They are often production operators or The Essence of Lean team leaders in companies with great pedigrees. Though they possess relevant knowledge and expertise, what they know best is how to carry out a lean operation as part of an integrated process.

Very few know how to set about modifying a value stream or business system to enable a new way of working, or have experience of putting in place the infrastructure to support and sustain a business system. The idea of recruiting from lean companies is part of a now widespread tendency to approach operations improvement from the bottom up. This may be a reaction to the laborious top-down programmes of 10 or 15 years ago such as TQM total quality management , which began with the CEO and filtered slowly down through the hierarchy to the rest of the organisation.

Starting from the shop floor probably seems like a better bet to managers tired of waiting for real bottom-line business benefits to flow from these top-down programmes. Imagine you reconfigure a production line to minimise the space it takes up. You may then be able to hold only a very small amount of line-side stock. Failures such as this all stems from a lack of understanding of what the transformation from a traditional operating system to a lean operation actually involves. True lean programmes challenge the organisational structures and formal management processes that surround operations.

They seek to diagnose and tackle cultural and behavioural barriers to change both on the shop floor and in the management suite. Imagine that a supermarket sets up a team to improve the way it goes about filling shelves so that it can free up more time for serving customers. What will it take to create sustainable change? In terms of the operation itself, what is needed is a clearly defined shelffilling process, supported by simple standards and practical training. In terms of management infrastructure, the team may decide to add visual tools such as a magnetic board with coloured counters that tracks who is working on which tasks and acts as a guide to decisions on labour allocation.

A few key 23 24 Journey to Lean indicators — say, cases handled per hour and time taken to unload a delivery — can be tracked on a daily basis and discussed during a brief team meeting every morning to review performance and agree priorities for the day ahead. Finally, staff will need to be engaged with the effort from the start so that they understand the reasons behind it and know what will happen and what roles they will be expected to play.

The way that people think and act and the systems within which they work need to change at the same time. As we have shown in this chapter, our approach to lean transformation is to create a stable change platform by working on all three elements of the business simultaneously: A failure to take account of any one of these is not just inadvisable but impossible, since working on one dimension will have a knock-on effect on the other two, leading to an unstable change platform.

5 editions of this work

In the next three chapters, we explore these three features of the lean landscape in more detail. The results had been great at first, but people had gradually reverted to their old way of doing things and the benefits had slipped away.

This is a pattern we see repeated all too often. The effectiveness of the lean operating system comes from the integrated nature of its practices and methods. Yet many organisations assume they can accomplish a lean transformation merely by applying a few lean tools. Such attempts are doomed to failure because there is no lean operating system in place to coordinate and drive improvements. One company we know devoted its efforts to changeover reduction SMED 1 across its entire operation; another sought to apply the principles of workplace organisation 5S as the basis for its improvement effort.

The problem with such approaches is twofold. First, lean tools are being applied without reference to business needs. Second, they are being used outside the context of a coherent operating system that can enable the organisation to meet these needs. For most companies, their operating system evolves haphazardly as their business grows. Imagine an entrepreneur who sets up a small bakery. She has few assets just her premises, a mixing machine and an oven and few resources or people to organise. Without realising it, she 1 Single Minute Exchange of Dies.

Things go well, and after a while her bankers lend her the money to buy a run-down bakery. A few years later, the business has developed into a medium-sized business with several sites, a range of products, thousands of customers and hundreds of employees. Whereas configuring assets and resources was once a simple matter there are only so many ways to arrange an oven and a mixer , it has now become quite complicated. Once the woman received a few phone calls in the afternoon that enabled her to plan her production for the following morning; now she relies on other people and an IT system to tell her what she needs to make.

Before, she had just one van that she drove herself; today she has a fleet of trucks that deliver fresh bread and other products to a range of outlets. For the most part, the configuration of assets, resources and people just happens; it is by accident rather than design. Lean by Design Our bakery example illustrated how a simple operation can be lean by accident; we now turn to a complex operation that became lean by design. We have already touched on the Toyota Production System TPS , the first — and still the pre-eminent — example of lean. But it does deserve further attention, for two reasons.

First, it has stood the test of time, having delivered superior performance over almost half a century. Second, it offers useful lessons for companies operating in sectors other than automotive manufacture. The Toyota Production System has three key elements: Just-In-Time Production The objective of Just-in-Time production is to produce and transport just The Lean Operating System what is needed, just when it is needed, in just the amount needed, within the shortest possible lead time.

This is a risky strategy. It incurs an inventory-holding cost and a consequent risk of obsolescence; it also fails to display true responsiveness to customer demand. Far from embracing lean, these companies are missing the whole point, substantially increasing waste rather than minimising it. True Just-in-Time capability is achieved when the products delivered have actually been manufactured in response to customer demand, not just delivered from stock. The process holds the smallest amount of stock needed to meet the delivery lead times customers require, thereby minimising the waste associated with inventory, reducing the risk of obsolescence, and providing a much more responsive system.

In practice, inventory needs to be seen as the lubricant needed to ensure that the system keeps flowing. These building blocks depend on the foundation of levelled production, which smoothes the workload over time. The reality is that variable work content — or differences in the total amount of time it takes to build different products — makes this impossible and demands a compromise, which is achieved through production levelling.

Most manufacturers would group similar products into batches to capitalise on economies of scale; Toyota does the opposite. The reason is apparent from the left-hand graph in the figure. Creating a batch of product L will mean that the production line is under-loaded; conversely, 27 28 Journey to Lean producing a batch of product H will mean that people have to struggle to cope with the extra work.

Such a method relies on standardised work and a highly skilled and adaptable workforce. To achieve this quality of workforce is no easy task, and attests to the crucial role played by management infrastructure in supporting the Toyota Production System. Imagine that two products, A and B, are made on the same line. This leads to a minimum average stock level of roughly half a week. Now imagine that production runs are reduced to a day rather than a week, as we see on the right-hand side of the exhibit.

This represents a reduction in average inventory of about two-thirds. However, this outcome does depend on customer demand being steady. If there are spikes in the demand profile, some sort of buffer stock will be required. The Lean Operating System Figure 3. Within a machining shop, for example, lathes and mills are housed in two separate groups. Although such an arrangement makes it possible for one person to operate several similar machines, the product tends to collect and stagnate between processes. This blocks material flow, which causes uncontrolled work in progress WIP to accumulate and prolongs lead times.

It also hinders communication between processes and increases the risk of quality problems that affect whole batches, since defects will be discovered only when the new batch is used in the downstream process. Wherever it is safe and feasible to do so, processes are 29 Journey to Lean 30 located next to each other in the actual sequence of operation. An operator can then work on one piece at a time and see it through to completion, instead of processing a batch of many pieces in parallel. As well as being more motivating for the operator, this arrangement minimises inventory and lead times and makes it easier to reallocate people from one product to another.

Continuous flow processing — often in a parallel or U-shaped work cell — is inherently more flexible, more visual, and more efficient, since it eliminates unnecessary movement and enhances communication. These features also make managing the production team easier. Takt time is defined as the total available time for production divided by the total customer demand for that period. Imagine an insurance company that is open for 50 weeks a year and In this case, the Takt time is 90 seconds total available time of minutes divided by 75 Takt time is designed to optimise material flow in pursuit of Just-in-Time delivery to the customer.

To do this, it sets the pace of production at the rate of demand, thus eliminating the risk of overproduction, which Toyota Figure 3.

Making Operational Change Stick

Once an operation is working to a steady pace, it is much easier for management to monitor performance, allocate the labour force and plan capacity. Takt time is used to balance work content in a continuous flow line, which often has the effect of reducing the amount of labour needed to build a product.

The normal approach is to balance all but one of the workstations to Takt, leaving the remaining workstation Operation 4 in the figure with a lower work content. In the short term, this spare capacity provides flexibility for dealing with any problems that occur during the production cycle. However, in the longer term, this remaining fraction of a complete work cycle becomes the focus for continuous improvement activity.

In many industries, customer demand varies significantly from period to period, so companies have to be able to cope with fluctuations. Although they can recalculate Takt time in each period to ensure that customer demand is always met, this will involve rebalancing lines and retraining people — a massive task in complex industries. Even Toyota confines itself to revising Takt no more than twice a year to ensure that changes are manageable.

This greatly reduces the need for central production planning. It also does away with the associated problems of reconciling stock and production data on an ongoing basis in order to maintain the integrity of the system. Many of us experience the benefits of a pull system in everyday life without realising it. The small controlled stock of burgers in a fast-food restaurant is an obvious example. This stock is determined by parameters such as the rate of production, level of demand and agreed safety stock, and is fixed at a level that ensures customers receive fast service and the risk of food wastage is minimised.

As a product is served to a customer, an order is generated to replace it with a fresh product. Pharmacies manage their stocks of drugs in much the same way. A visitor from the company noticed that the retailer was replenishing only those products that had been taken off the shelves by customers — a method it had implemented to reduce wastage in perishable items. The Toyota employee recognised that such an approach could easily be adapted to reduce waste in a manufacturing environment.

There are many ways of applying pull systems, but the essence is that parts are produced only once a signal — a kanban — has been received from the downstream process. The exact configuration of the system will depend on a number of variables, such as the size and value of the components; the number of variants and level of demand for each one; the information lead times; and the shelf-life of the product. Working in this way has the general effect of increasing the frequency and number of changeovers.

This is why changeover reduction or SMED techniques form an intrinsic part of lean. Indeed, in many cases, it is reduced. A good rule of thumb is that roughly 10 per cent of total processing time should be spent on changeovers. Toyota uses pull systems to target the most serious form of waste: Surplus inventory masks and compounds other forms of waste. Removing it puts equipment reliability at a premium. It is for this reason that approaches used to maximise the reliability and utilisation of equipment in a systematic way, such as Total Productive Maintenance TPM and Reliability-Centred Maintenance RCM , are an intrinsic part of lean operating systems.

Examples can be found in the aerospace, biomedical, electronics and process industries. Autonomation Autonomation is a process designed to allow a workforce to detect production problems quickly and resolve them decisively. Its objectives are to improve equipment reliability, enhance product quality and increase The Lean Operating System productivity. It consists of three elements: Toyota found that the most effective method of detecting a fault or abnormality is to give the responsibility to the people who operate the process.

They must have a good understanding of what the customer requires so that they can decide when a product or part has gone beyond the acceptable limits of the process. To help operators detect problems, Toyota employs visual standards and mistake-proofing or Poka-yoke techniques. Poka-yoke techniques are used to prevent abnormalities occurring in the first place, and to stop a process automatically when an abnormality is identified.

An everyday example can be found in cars with an automatic gearbox: Another means of detection is to monitor the production rate and compare the actual rate to the target, which is based on Takt time. Once problems have been detected, the process should wherever possible be stopped. This can be done manually or — in cases where the stop mechanism is linked to the detection process — automatically.

In process industries, it may not always be feasible to stop the process, but it is still important to detect the problem as quickly as possible so that corrective action can be taken. Such industries often use statistical process control to monitor the processes and detect when a problem has occurred. Once a problem has been detected, the process or the operator needs to alert the team leader.

This can be done verbally, by means of an Andon board a display that indicates production status , or via an audible alert. It may be, for example, that an additional check needs to be made on every component until the source of an intermittent fault upstream in the process has been identified and addressed.

Journey to Lean: Making Operational Change Stick - PDF Free Download

Flexible Staffing Systems As Toyota strives to match its production to true customer demand, a point will come where the variation in demand for individual products will mean that certain flow lines are under utilised while others are overburdened. At this point, managers will need to reallocate resources quickly as demand changes. The mechanism that enables them to do so 33 34 Journey to Lean is known as the flexible staffing system. Its objective is continuously to optimise labour productivity to whatever the level of demand may be.

In an assembly environment where the demand for different product lines is relatively predictable from one week to another, this is achieved by changing the Takt time of the lines involved and then dividing the total work content by the Takt time to determine the number of people required at this level of demand. In practice, the tasks an individual operator needs to perform may well vary as the Takt time changes.

This is why standards such as standard work sequence charts are needed to define the precise role each person must play at a given level of demand. Consider a manufacturing area that has two cells, A and B, building two different product groups. The processes are well documented, with standards for each task that are used as the basis for training. Operators are trained in both sets of processes so that they are able to work on either line. The team leaders have worked out a line balance for three levels of demand: They work with the production scheduling team to understand demand changes and agree Takt times on a week-by-week basis.

Journey to Lean

There is no need to build excess stock to keep people busy in quiet periods, or to run overtime at peaks in demand. This simple example also shows how standards underpin an effective operating system and are central to any flexible staffing system. Clearly defined work standards and a good understanding of the work content of each job are prerequisites if people are to be moved between cells to rebalance lines without creating confusion. This is one of the reasons why Toyota and other Japanese manufacturers emphasize the need to standardise operations and so create a foundation for flexibility.

One important manifestation of standardisation is the use of visual management to make the status of an operating facility obvious to management and workforce alike. This extends from the overall layout of a process within a site, which should follow a logical flow, through to the storage of frequently used tools on shadow boards. Such transparency makes it possible to give front-line workers the The Lean Operating System Figure 3. Employing visual management also engages the right side of the brain, which processes images and shapes holistically rather than analytically.

People can see whether something has gone wrong much more quickly than they could diagnose it intellectually. Having examined the key elements of the Toyota Production System, we now turn to the organising principles that characterise the lean operating system. Any would-be lean company needs to design an operating system that takes account of its own particular context: However, we have identified seven key imperatives that serve to sum up the distinguishing features of a lean operating system.

In reality, a flow may be so haphazard and disconnected that the very notion of a value stream may appear meaningless. The first principle of a lean operating system, therefore, is that it is based on value streams that are created by grouping together similar products. Exactly how the products should be grouped will depend on the business, but the criteria are likely to include patterns of demand, product characteristics and process routings. In many cases, the product and process criteria will define the value streams relatively easily.

One manufacturer that made shock absorbers had two value streams based around two different technologies that required different equipment for the downstream half of the process. In other cases, the product groupings will be less clear-cut. At one cosmetics factory, the operating system was reconfigured to create one value stream dedicated to steady products even though there were many of them and another for promotional products, whose demand characteristics were completely different. At Airbus UK, the value streams were defined as single-aisle, long-range and wide-body aircraft.

In a service industry, the task of identifying value streams calls for creative thinking.

In a hotel, for example, value streams might be The Lean Operating System defined in terms of customer needs such as rest, food and beverages, and recreation. Such a grouping might then lead to changes in the design of processes and allocation of resources. It might be that the logistics supporting rest such as laundry services would be treated separately from the logistics supporting food and beverages such as the ordering and delivery of meat.

Once it had established the value streams in this way, it determined the expected volumes that would be handled by each one. Flow the Value Along the Stream from Beginning to End Having established its value streams, a business then needs to design its operating system so as to ensure there is a flow of value to the customer from the start of the value stream to its end. The value stream should be configured to enable this product to be ordered, manufactured and delivered to the customer within the shortest possible lead time.

Whenever the value stream is broken, the execution of production processes is delayed and inventory increases. Though sharing equipment is sometimes unavoidable as with especially expensive or complex machinery , it is more often the result of misguided attempts to cut unit cost, as opposed to optimising total cost.

Sharing equipment carries costs of its own, as attested by the bottlenecks that emerge in any process using shared resources. Although dedicating equipment and resources to a single value stream may require a net increase in investment, it is often more cost-effective in the long run once the benefits of lower inventory, shorter lead times and fewer quality risks have been added to the equation.

Implicit in the principle of maintaining flow is the objective of minimising the leakages that can occur along the value stream. The challenge is to build an effective process through which value is added and leaks are eliminated.

Account Options

Consider the leakage caused by excess inventory in the value stream. It could be that operators are stockpiling inventory to compensate for the vagaries of an unreliable machine. If the machine is made much more reliable, they will no longer need to build up these excess stocks, and the leak will have been plugged. By instructing each process what to do next, it drives the business system to ensure that the right product or service arrives at the right place at the right time. It includes the allocation of people to value streams and the way they use equipment within a value stream.

It is necessary, though often difficult, to optimise all three flows at once. Consider a telecommunications operator working to improve its process for repairing faulty phone lines. Complaints received at the call centre must be passed on to the teams who diagnose faults, plan the work and make the repairs. Ensuring the timely flow of this information and maintaining the integrity of the data is central to creating an effective end-to-end process.

The workload and routing of the field technicians who make the repairs — the people flow — is a key determinant of both the lead time to repair the fault and the productivity of the workforce. Materials must also flow efficiently to ensure that the field team always has the tools and components it needs to do the repairs and yet the business is not burdened with unnecessary stocks.

Pull Products at the Points where the Flow must be Broken In some sectors, such as retail and assembly, it is inevitable that the flow of value will be broken at some points in the operation. Perhaps a process needs to make use of an expensive piece of equipment such as a paint plant that has to be shared between several value streams. Or perhaps the value stream is organised not around predetermined processes but around highly variable interactions between people, such as in renting a car. Such situations produce a less precise value stream whose flow is susceptible to random interruptions.

The Lean Operating System Imagine a painting process in which parts required by a number of downstream assembly lines are painted in batches in different colours. But we can minimise disruptions to the flow by setting up a simple pull system. It involves storing a defined quantity of painted parts after the paint process so that the assembly lines can take parts as they need them. Whenever they take a box of parts, a kanban attached to it is returned to the paint process.

Once a predetermined number of these kanban have accumulated, the paint process changes over to paint parts to replace those that have been used up. This regulating mechanism is a simple and efficient way to determine when the process should change over, what it should make and in what quantities. It allows the process to maintain the supply of parts while minimising the stock that is held.

Flex the Operation to Match Customer Demand The first three principles of lean operations have dealt with the broad theme of configuring assets and resources to create material flow. There remains the overriding question of how best to introduce the demands of the customer into the value stream in the first place. This is where the best lean operators excel and the weaker ones fall short. A lean operator is able to do everything it takes to satisfy a customer without carrying out activities or introducing product features that add cost but no value.

This requires an understanding of true customer requirements and a capability to flex the scale and pace of operations in order to follow the customer. Just as the pull principle often translates into the tool of kanban, the principle of flexing an operation to match customer demand often translates into the concept of Takt time. Suppose a private hospital specialising in a particular operation has demand for 16 operations a day and that the operating theatre is open for 8 hours a day, then this gives a Takt time of 30 minutes.

This means that assets operating theatre, beds, equipment, surgical tools and resources doctors, nurses, administrators should be organised and optimised to complete one operation every 30 minutes. The Takt time will drop to 20 minutes, and assets and resources will need to be reconfigured and probably augmented in order to achieve it.

In a chemicals plant, for instance, the product mix needs to be managed to match demand as closely as possible even where the physical rate of production is determined by chemical processes or equipment considerations. Strongly seasonal markets present a different kind of challenge. Take an ice-cream manufacturer. Its operating system must find ways of flexing production to accommodate the ebb and flow of customer demand. It may take on temporary workers at peak times, or enable its workforce to bank additional hours in the summer months that they can then take as holiday during the winter.

At a retail bank, the operating system will need to accommodate peaks in demand at lunchtimes. This will probably mean that multi-skilled members of staff will need to play a customer-facing role during these peak hours and revert to a back-office role at quieter times.

Introduce Information Defining Customer Requirements at a Single Point, and as Late as Possible in the Process Managing operations through a centralised planning or scheduling function is much like managing a centrally controlled economy. It should work in theory, but in practice it seldom does because of the difficulty of managing variations in a timely fashion. Centralised systems rely on a few basic underlying assumptions.

Once these are in any way challenged because a delivery is late, or a batch of parts is defective, or a machine breaks down , the system quickly deteriorates, leading to a vicious cycle in which scheduling grows increasingly dependent on individual interventions instead of the central system.

Consider a simple manufacturing process that makes metal components for lawnmowers in three steps: Suppose that customer demand for a given day is identical parts. A typical production control system would translate this requirement into a build list for each process. Suppose that, historically, pressing has operated with an average defect rate of 5 per cent, welding 10 per cent and painting 20 per cent. The production control system would issue an instruction to pressing to make parts to ensure that make it through to the customer divided by the cumulative defect rate, 0.

The Lean Operating System However, the one thing you can always rely on is variability. On any given day, the defect rates may not be 5, 10 and 20 per cent, but perhaps 10, 10 and 25 per cent. In this example, only parts would make it through to the end of the process: To increase the production target is not the answer; on another day with a different pattern of defect rates, starting with the requisite number of parts could lead to a substantial surplus. As this example illustrates, traditional production control systems tend to offset the risk of failing to meet customer demand by increasing batch sizes and adding buffer stocks.

This adds cost and increases lead times. The complex conditions surrounding most production processes can only magnify the problem. As well as managing defect rates, the operating system must also take account of assumptions about such things as machine breakdowns and changeover times.

As product and process complexity increases, the challenge of managing information flows gets steeper. So how would a lean operating system differ? It seeks to introduce the information defining customer requirements into the system in such a way as to avoid the overproduction and shortages that plague centrally planned economies and mass producers alike. The key is to introduce the information at a single point local to the value stream, rather than at every point in the value stream or at some point remote from it, and then hard-wire all other processes to this information at the point of introduction, thereby establishing a physical connection between production and demand.

This is possible because the Takt time governing the production rate is itself determined by actual demand. Once the information defining customer requirements has been introduced, the system pulls the product or service through the value stream accordingly. The precise mechanism it uses will depend on whether operational lead times are longer or shorter than customer lead times. If shorter, products can be made to order; if longer, they must be made to stock.

Suppose that manufacturing lead times for lawnmower components are longer than customer lead times. In a lean system, we would need to hold a defined stock of finished products after the paint process. The withdrawal of these parts triggers a 41 42 Journey to Lean kanban to be sent to the paint process telling it to take parts from the stock of unpainted parts and paint them to replace the components that have been sent to the customer, and so on up the value stream. This method of production control eliminates the need to second-guess all sorts of uncertainties.

If 20 parts should need to be scrapped from the paint batch of , then 20 replacement parts will be taken from the upstream shop stock, generating a signal to press an additional 20 parts to replenish this stock. In other words, the actual and not the anticipated demand is communicated up the value stream. Consider the way ball boys and ball girls deal with variability on the tennis courts at Wimbledon. Having been trained, they are given responsibility to make decisions for themselves rather than refer to authority.

They are able to go about their business in the company of highly paid international tennis stars without distracting them or getting in the way. Without this freedom of interpretation, standardisation can indeed become constricting. Operating standards ensure that the safest and most efficient way of working is defined and repeated. This has benefits for customers, who will see better and more consistent quality; shareholders, who will gain from higher productivity; and employees, who have clear and safe procedures to follow. Standards also greatly reduce the risks associated with introducing new products or changing a process.

At the most basic level, standards ensure that tasks are done the same way no matter who is doing them. They also provide a foundation for training, and a baseline for improvement activity. They should be seen as living documents, constantly updated by work teams as they apply The Lean Operating System them. As processes are improved, standards need to be updated to capture the changed conditions, which then become a new baseline for improvement.

Whether they are to do with greeting a guest in a hotel, answering the telephone in a call centre, or building a car, standards assure the quality of an operation and thus protect the brand and the market premium it commands. They also play an important part in flexible labour systems by making it easier for individual employees to switch from one task or work cell to another. This enables companies to respond quickly to fluctuations in customer demand and thus maximise their productivity.

The use of visual management in the workplace also plays a vital role in building flexibility within the operating system. As well as making the location of materials and tools transparent, it can be used to create commonality between different work areas and between different sites, making it far easier to transfer people from one workplace to another. In such a capital-intensive environment, the cost of not producing is extremely high, but Toyota has judged that the cost of producing defective products is higher still.

This is unusual; in most other companies, defects are handled in one of two ways. Either a group of people with special responsibility for quality is called in to investigate abnormalities and decide what to do, or else the news that a quality problem has occurred makes its way up the hierarchy until it reaches someone with the responsibility or courage to do something about it. In both cases, there is a substantial time lag between the detection of the problem and its resolution, during which the production of defective parts may well continue.

Contrast this with a lean operating system. Abnormalities are detected as soon as they occur, and if possible the process is stopped and not allowed to restart until the root cause of the problem has been identified and contained. This forces the organisation to institutionalise its problemsolving capabilities and pushes it towards continuous improvement. Such 43 Journey to Lean 44 a rigorous process benefits customers by weeding out defective products and services, shortening lead times and reducing costs.

Putting it All Together Because lean takes a holistic approach to the design of an operating system, it ensures that the root causes of underperformance, and not just the symptoms, are identified and treated. A business seeking to identify sources of loss will need to examine the whole operating system, from customers all the way back through the manufacturing processes to suppliers.

To understand this better, consider the case of a small sheet-metal fabricator that made parts for custom-built air-conditioning units. In an attempt to improve performance, its operations manager had already removed one bottleneck by replacing two old punches with a highercapacity and more flexible new CNC punch.

The process he was working on involved cutting sets of parts according to a job list. However, the sequence was often disrupted by a lack of the appropriate grade of metal. After being cut, parts were stored and then moved to a workshop containing three folding machines. Although these folders had a sequence list to follow, they tended to cherry-pick jobs to suit the set-up of their machines. Deliveries to customers were consistently late. The operations manager believed that the folding machines represented a bottleneck, so he ran some workshops to cut changeover times and help the folders work more closely to the sequence list.

He was surprised to find this made no difference to delivery performance. Indeed, the folders often stood idle for long periods. When the company took a step back and looked at the process in a more holistic way, from the supply of steel through to delivery to the customer, a different set of problems emerged. These related to the flow of information and the replenishment of sheet steel — the true sources of loss. Another manufacturer committed itself to constructing an expensive new workshop when it ran out of space. But a systems perspective revealed that most of the existing floor space was taken up by inventory.

The weak points in the operating system turned out to be poor information flows as much as bottlenecks in material flows. Tackling them allowed the company to reduce inventory drastically within a few months. Had it done so earlier, The Lean Operating System it might never have needed a new building, and could have avoided a hefty investment. The first step in transforming operations is to design and develop an effective operating system: The next chapter sets out our approach to developing a management infrastructure capable of supporting and sustaining the new operating system.

Operational performance improvement programmes have earned a bad reputation. They tend to fail. Many deliver temporary gains, but few succeed in sustaining the early benefits, and fewer still manage to establish a genuine culture of continuous improvement. In most cases, operational performance either slips back to its old level or stagnates within a year. The latest initiative is soon consigned to history, joining a host of previous efforts in the company archives. Though some changes are irreversible, such as process automation or the redesign of physical layouts, most are not.

One defence manufacturer launched a pilot improvement project in its machine shop, which makes large and complex parts for later assembly. Senior managers believed that the machine shop was running out of capacity and acting as a bottleneck in the manufacturing process. The primary objective of the pilot was to improve the equipment utilisation the overall equipment effectiveness, or OEE of a number of machines, each of which had a capital value of several million euros.

The manager in charge of the machine shop believed that raising the OEE would enable him to release substantial processing capacity and thus 47 48 Journey to Lean accommodate future business growth without major capital investment. A project team identified ways to eliminate problems and developed ideas for technical improvement, such as moving stocks of raw materials and improving loading procedures to eliminate unnecessary machine and crane operations. When implemented, these ideas quickly led to an improvement in OEE of over 50 per cent. Within a few weeks, however, the improvements began to fall away, and after six months OEE was only about 20 per cent higher than when the initiative began.

Closer examination revealed that his span of responsibility was too great: Operating such a flat organisation left him feeling that he was trying to hold on to the gains single-handed. Such a structure might just have worked in a stable manufacturing process where operators performed unskilled tasks repetitively, but it was doomed to fail in a relatively unstable and complex process demanding highly skilled operators. The company decided to add a new layer of four shift managers, each leading four or five team leaders, to enable it to recapture and sustain the improvements it had achieved in the pilot.

Every company undertaking a lean programme will need to consider whether its formal management processes, organisation structure and capability-building infrastructure reinforce its new lean operating system. If the management infrastructure is not aligned properly, the operation is unlikely to achieve its performance objectives. To see how this works, consider a company that has introduced new kanban arrangements for organising in-process inventory or parts stocks.

Properly implemented, these arrangements will reduce stock-outs, smooth flow, simplify production planning and enable Just-in-Time delivery. But operators and production planners can only implement them correctly if other processes and structures are brought into line. Planning and scheduling will need to be revised to prevent competing production instructions being issued to upstream processes.

New local performance metrics and management review processes must be introduced to provide incentives for operators to respond to the kanban. Training will be required, and may entail changes to skills matrices and records. So how does a company go about aligning its management infrastructure with its operating system? In the rest of this chapter we describe the key elements in developing a lean operating environment.

Management Infrastructure We have identified five key elements of a management infrastructure designed to support a lean operation see Figure 4. The relative importance of each element will depend on the operational setting. We have found that three key aspects of organisational structure determine whether it is aligned with the operating system: Team Size At Toyota, a typical production line team consists of between four and eight employees with a working team leader. Why do team sizes vary so much, and does it matter? Size matters because getting it wrong can be disastrous: Sometimes front-line teams are structured, sized and skilled with considerable thought.

More often, they evolve through custom and practice until the rationale for their structure and size has been lost and they may no longer be geared to the demands of the technical processes they operate. So how should lean companies determine the size of their front-line teams?

We have identified several criteria that we believe managers should consider when determining optimal team size see Figure 4. These criteria apply to service industries and back-office processes as much as to manufacturing industries. Conversely, the more stable the process, the fewer interventions managers need make and the wider the management spans can be, allowing larger front-line teams. The more critical the process is in terms of the cost of any malfunction, the more time the manager or team leader will need to spend dealing Management Infrastructure with problems when they arise.

In such situations, a smaller span of management reduces the risk of expensive failures. One well-known consumer goods company is able to have large teams not just because most operators carry out simple, repetitive packing tasks, but also because it employs first-rate and well paid shift managers.

Countless different models exist for defining the role of the team leader or front-line manager. They vary in the way they group together process steps, organise work shifts, assign roles between teams and their managers, and so on. At one extreme is a self-managed team that runs operations autonomously by dividing up responsibilities between its members.

At the other is a front-line manager who bears full responsibility for a line or shift, and combines the duties of a supervisory team leader with broader aspects of HR management such as recruitment, formal appraisal and discipline. Between these extremes lie a multitude of other models. A working team leader might combine direct operational tasks such as covering for absentees with some day-to-day personnel management such as managing overtime. A supervisory team leader, on the other hand, might spend all of his or her time managing the hour-by-hour needs of the process and the operators under his or her control, doing no operational work and holding no responsibility for annual staff appraisals or discipline.

So which model works best? Everything depends on the context. At one manufacturing company we know, self-managed teams work well because the nature of the process largely determines where operators are located and the tasks they perform, and because many of these operators have years of experience. About half their time is spent covering for absentees, about 20 per cent is set aside for responding to problems highlighted by Andon calls, and the rest is devoted to tasks such as training or off-line quality checks. Both production roles and team sizes are clearly defined.

The company works to a base level of 5. In an area such as assembly where considerable on-the-job training is needed, for instance, teams will be smaller to give the team leader time to provide it. In highly automated areas such as the press shop, teams can be larger. Spans of Control and Levels of Hierarchy The criteria used in determining the size of front-line teams also apply to the design of the management hierarchy.

Managers running complex departments with unstable processes or poorly trained subordinates need small spans of control. Those in charge of stable, repetitive operations with competent and well trained subordinates can cope with much bigger spans. As for the levels of hierarchy, they should be kept to the smallest number consistent with appropriate spans of control at each management level.

Where levels are few, management is physically closer to the shop floor and feedback times are shorter, improving information flow and clarity and speeding up decision making. Even in a large operation, it should rarely be necessary to have more than three levels between, say, a team leader and an operations director.

In manufacturing, the roles corresponding to these levels might be shift manager, area manager or superintendent, and production or factory manager. Although minimising the levels of hierarchy may limit opportunities for promotion and management development, it has the virtue of simplicity. It helps everyone understand their roles and enables senior managers to organise their management and communications processes effectively. This calls for a good understanding of the relationship between mindsets and behaviours as we shall see in Chapter 5 as well as effective performance management processes.

In our Management Infrastructure experience, the performance management system is often the weakest link in the management infrastructure that underpins the lean operating system. One business we know took pride in its performance reporting system, which was capable of producing any report you might think of at the touch of a button. Fortunately, the problem was fairly easy to correct, but this example does show how the mere presence of a reporting system can delude management into assuming it is in touch with performance. One fast-food company used food scrap performance as a criterion for evaluating and rewarding store managers.

Though this approach succeeded in focusing managers on an important cost driver, customers often faced a long wait for their orders. A great performance management system in a lean operation is much more than just a colourful set of reports, or the tracking of the right metrics; it goes right to the heart of management. It calls for not only a clear system definition — the right metrics, supported by effective tracking and reporting processes, IT tools and linkages to financial and other systems — but also the right approach to managing the dynamics of performance hour by hour, shift by shift, day by day, month by month.

Setting up a lean performance management system involves several steps: Designing the System Before a company can start actively managing performance, it needs to establish the right formal architecture for its operating system. It should begin by articulating its top-level business goals: Selecting these key operational objectives is not as simple as it sounds, and companies often have too many; we worked with one company that had 21!

Making trade-offs across so many objectives on a daily basis is impractical and produces unpredictable results. Instead, managers need to be presented with a concise list of indicators to help them prioritise. The next step is to build a hierarchy of metrics throughout the organisation that link the main drivers at the lowest levels back up to the core toplevel objectives.

Care is needed to ensure that the lower-level metrics add up to those at the higher level, that all key metrics are covered at each level, and that each metric has a clear mathematically based definition so that the data required to calculate it is readily available or can easily be obtained through changes to data-capture mechanisms. Once these key performance indicators KPIs have been defined at every level, managers must decide how they will be reported and who will be responsible for collecting and reporting them.

If an IT tool is to be used, it will need to be evaluated to ensure it is fit for the purpose or capable of having its database architecture or algorithms modified. Often a new KPI definition will require changes to the manual data collection methods that feed into the IT system. Setting Targets This step in the process links the one-off system set-up with the dynamic performance review cycle. The top-level objectives must be carefully reconciled with the bottom-up improvement plans developed by the frontline teams and their leaders if the improvement plans are to be robust.

Ambitious stretch targets should be treated with caution. They can prove an effective shot in the arm for businesses that need to make a step-change to survive and for those trying to galvanise a workforce into making bold changes; used routinely, though, they soon become demoralising. People quickly become worn down and lose heart if they are set targets they know they have little chance of attaining.

Monitoring Day-to-Day Activities Needless to say, this must take place at every level of the organisation, though it will probably take different forms and occur at different frequencies. A team leader might well cycle around the shop floor once an hour; a chief executive, once a quarter. John rated it really liked it Feb 10, Oremeyi Akah rated it did not like it Sep 05, Louise added it Mar 09, Vicky marked it as to-read Oct 19, Anthony added it Jan 16, Pawanraj marked it as to-read Feb 22, Emil F marked it as to-read Jan 18, Ryan Lilly added it Mar 02, Robert Gulbahce marked it as to-read May 06, Malakia Mafuma marked it as to-read Sep 12, Pearlz is currently reading it Dec 05, Mark marked it as to-read Apr 28, Lukasz added it Feb 07, Rodrigo Serrano added it May 10, Candice marked it as to-read Jun 26, Ryan Greaves is currently reading it Sep 03, Alan marked it as to-read Sep 13, Paco Lendinez is currently reading it Nov 08, Genco Ilgin marked it as to-read Mar 13, CCI added it Mar 24, Joris added it Jun 28, Tunde marked it as to-read Jul 22, James Towers marked it as to-read Jul 27, TelfordASE marked it as to-read Sep 13, Anirban Majumdar added it Oct 18, Mikael Rask added it Oct 27, Isar Soleimani added it Nov 20, Manuel Morales marked it as to-read Dec 01, Andrew Bishop marked it as to-read Dec 25, Daria Gan marked it as to-read Mar 01, Lakna marked it as to-read Mar 19, Anan Wiwhet marked it as to-read Apr 02, There are no discussion topics on this book yet.

Books by John Drew.

• How To Create, Protect and Promote Your 5 Star Reputation Online
• Orchestrating Inquiry Learning
• Mere Christian Apologetics
• The Cabo Contract A Colin Pearce Adventure (Colin Pearce Series Book 2)
• Edens Jester
• The Sound of Death (Miranda Sykes Series Book 2)
• Wir sind die Neudorfer, Band 2 (German Edition)