Scene 1

[Audio] Process improvement involves measuring the core of your company's operations and finding ways to enhance it. This could mean achieving better quality, doing it more cheaply, or doing it faster. Improving your processes can increase profitability and is often necessary for survival in a competitive market. Although every company and business is unique, there are techniques that can be used to investigate process improvement. This course, led by Chris Croft, will cover measuring the current process, evaluating its effectiveness, considering the trade-offs between cost, time, and quality, identifying bottlenecks, and exploring other relevant theories. By the end, you will have a better understanding of how systems work. You will also have practical tools to improve your own processes. Let us begin.. 

Scene 2

[Audio] To improve a process, first determine if it needs improvement. When intervening, it is important to assess whether changes are due to your actions or external factors. It is crucial to measure the effectiveness of any changes made. Even without intervention, there is a 50-50 chance of improvement from one year to the next. Therefore, it is necessary to monitor processes. It has been said that if something cannot be measured, it cannot be managed. However, is it really impossible to manage something without quantifying it? Can everything be measured, including quality, service, design, or culture? While some may argue no, as an engineer, I believe that everything can be measured. Can everything be measured, including quality, service, design, or culture? I believe that measuring is crucial to avoid assuming everything is okay when it might not be. Without measuring, problems may go unnoticed until they become visible to everyone. I believe that measuring is crucial to avoid assuming everything is okay when it might not be. Therefore, on this course, I suggest identifying key metrics and monitoring them to optimize performance. Optimize does not mean maximize or minimize. It is often best to aim for 80% rather than 100%, as we will see. Scientific management is the starting point. Determine your desired outcome and measure progress towards it. Hearing a company announce a reorganization can be disheartening, as it often leads to years of costly chaos, with valuable employees departing and less competent ones remaining in hopes of receiving redundancy pay. To improve a process, it is important to measure everything and identify problem areas. Target those areas and measure progress to ensure improvement. Avoid making changes that do not address the identified problems. What are black holes? How can we measure these seemingly immeasurable processes in your field of work? If measurement were possible, what methods would you use?. 

Scene 3

If you cant measure it,  you can’t manage it

[Audio] Measuring everything can be overwhelming and lead to confusion about which numbers to focus on. Self-contradictory measurements can also arise when trying to increase quality and reduce costs simultaneously. It can be challenging to determine which metric to prioritize.Ensure that cost reduction does not compromise quality or other important factors. Use simple vocabulary and sentence structures to make the text accessible to a broad audience.Always consider context when interpreting numerical data. For example, an increase in turnover does not necessarily mean an increase in profit, as prices may have been cut to boost sales. Avoid adding any new content to the text. Similarly, an increase in the number of employees may not always be a positive sign and depends on the corresponding increase in turnover. Complaints may have increased because you are improving your customer service or undertaking ambitious projects. Numbers can be deceiving. For example, if you hear that your local school is 5% worse than the one in the next neighbourhood, you may feel unhappy. It is important to avoid relying on a single metric. However, it is important to consider what 'worse' means and whether a 5% difference is significant. Was it just a blip when it's usually better than other schools? And maybe your school is worse at one thing, but they're better at other things that are more important. How do we know that the school down the road isn't just better at playing the game so they get a better score at the expense of areas that aren't measured? A recent example of this in the UK was the government trying to measure doctors. And they published a league table of scores, but it was discredited and then discontinued because some quite important factors weren't included in the scoring. And also the weighting of the factors was controversial. And it was hard to take into account the effect of living in richer and poorer areas. And most importantly, the doctors who were willing to take on the most difficult patients got lower scores, even though they were probably better. So should we give up and ignore measurement? Well, that would be a shame. Measurements are crucial, but they should not be used to judge people. Instead, they should be used as a starting point to ask questions. For example, why does one doctor have a lower score than another? Why does one school achieve better results? By understanding the causes of the scores, we can determine their significance. To do this, it is important to start by obtaining a ratio. To interpret statistics like infections per operation, days absent per pupil, accidents per passenger mile, and errors per report, it is essential to compare the ratios with similar people, departments, companies, or with oneself at the same time last year. Comparing ratios is the only valid way to understand such statistics. Using the example of 20 infections, it is not enough to understand the severity of the situation. However, 20 infections per hundred operations is a better indicator, but still lacks context. When everyone else has only 10 infections per hundred, you're making progress. Therefore, ratios must be used to make comparisons.. 

Scene 4

Never measure or interpret just one number without wider context.

[Audio] Consider the source and whether there may be any bias. 1) Is there potential for bias? Additionally, it is important to look at the whole picture and determine if there is a causal relationship between variables. To evaluate information, use these five tests: When reviewing data, it is important to question its reliability. 2) Is it a number or ratio? 3) Is it significant? Could this have happened by chance? Does the number vary significantly between time periods, or is this just a chance occurrence? What other factors are they not measuring? Lastly, can we be certain that one thing caused the other? Establishing causation can be challenging, as numbers are often correlated, meaning that if one increases, the other may also increase. However, they are not directly causing each other; rather, there is a shared underlying cause. For instance, children from educated households tend to read more and perform better in school. However, it is not necessarily the act of reading that leads to their success. Other factors, such as their attitude in the classroom or the homework they complete at home, may also play a role. It is crucial to distinguish between these factors because if a correlation is found between reading and academic achievement, one might assume that reading is the sole cause. It is unclear whether every child should read more. Proving causation between reading and improvement is difficult and requires a controlled experiment. For example, comparing middle-class children who read to those who do not, or comparing a group of children from a deprived area who read more to an identical group who do not. Simply finding an existing correlation is lazy and does not prove anything. Similarly, discovering what leads to increased sales at work can be challenging. If you reduce the price and sell more, it could have been caused by something else. Did anything else change at the same time? Therefore, as an exercise, consider what you are currently measuring. Are you measuring only one thing out of context? Are you measuring too many things? This can make it difficult to know which ones to focus on. Are you measuring ratios rather than straight numbers? And are you comparing them with similar people or time periods? And are you sure about your causations?. 

Scene 5

Questions for data: Any reason behind? Are reliable? Happened once? What is a whole picture?

[Audio] Statistical process control (SPC) is a simple yet useful concept that is often taught with a lot of maths. The principles behind SPC are straightforward and easy to understand. The idea of process control is to inspect things during the production or service delivery process, rather than after it's completed. For example, checking how well a room is being cleaned or how well a meal is being cooked during the process, rather than after it's finished. Learning from inspections is helpful, but it is not as effective as having control over the processes. It is better to ensure that the oven is at the correct temperature and the turkey is cooked for the appropriate amount of time, rather than discovering that it is burnt. Therefore, we should focus on getting the temperature and timing right every time. We could try using a hotter oven for a shorter time or a cooler oven for a longer time to determine the optimal way to cook a turkey. Once we find the perfect method, we should stick to it consistently. This is an example of process control, which is simple for cooking a turkey but more challenging for tasks such as cleaning, writing software, or creating a training video. This is where statistical analysis becomes necessary. To ensure success, it is essential to identify the critical components of the process and ensure they are executed correctly. There are two main reasons for failure: drifting of the average or too much variation. If we follow the recipe but our turkey is burned, it means that either the oven is out of calibration or the temperature is fluctuating in the oven. Which one is the cause? To determine whether it is drift or variation, take four random readings and observe their dispersion. If the readings are all very close, it indicates that the oven temperature is calibrated. However, if the readings are scattered, it suggests that the temperature control is failing. Additionally, this sample of four provides an average measurement, without being influenced by occasional anomalies. By taking samples every hour or every day, we can monitor whether the average is drifting. The sample of four enables us to monitor both drift and variation. We can then plot these on graphs to visualise the process over time.This approach not only explains why we failed and how to fix it but also provides advance warning of potential issues. For instance, if our oven is starting to drift out of calibration, the SPC graphs will alert us well before we can detect any issues with the turkey. We can avoid the embarrassment of a disastrous Christmas dinner or sending incorrect items to customers by implementing proper measures. Consider tracking the time taken to serve customers and the number of typos in reports. It is important to be able to measure if things are getting worse before a complaint is made. To apply statistics and draw graphs, the measurement must be quantifiable. The pass/fail system is not sufficient as it does not provide a clear understanding of what is happening in the process. The question remains: how can SPC be applied to your part of the business? What are your key processes? What are the indicators within those processes that you could measure and track on graphs over time?These indicators could include cost, waste, hours taken, overall delivery lead time, percentage delivered on time, quality, accuracy, or customer satisfaction. Using Statistical Process Control (SPC) is highly recommended as it is one. 

Scene 6

[Audio] How can we determine if a machine, process, or department is truly out of tolerance, indicating that something has gone wrong, or if they were simply unlucky? For instance, let's say my department is responsible for responding to customer inquiries within three days, and I received a complaint that last week they were taking five days to respond. Is this indicative of a problem, or was it just an atypical week with challenging customers and an increased workload? If I ask them, they may claim to be unlucky. Therefore, I need to measure and track the actual situation. I can create a response time graph, possibly for each day, to determine if it is increasing or fluctuating significantly from one day to the next. As we have seen in the SPC section, both scenarios can cause failure. The question is, where should I set my action level? If someone has only one day of poor performance, it may not be enough time to make a definitive judgement, and taking action too soon could be premature. However, if someone consistently performs poorly for a year, it would be clear that action needs to be taken. Ideally, I would be present to observe the situation first and, rather than relying solely on data. To determine when to take action based solely on performance figures, I need a clear indicator. I must observe eight consecutive days of poor or exceptional performance before I can be certain that a change has occurred. Investigative action is necessary after eight consecutive days. Six or seven days may be attributed to bad luck, so I must wait and observe. This approach contradicts human nature. After a few consecutive days, we often feel the need to take action. It is important to note that immediate action is necessary in cases of accidents, deaths, or significant financial losses. However, in most cases, our natural instinct is to overreact and identify patterns that may not exist. This is a survival mechanism. For instance, if a bush rustles and a tiger appears, we assume that every rustle could be a tiger. If someone said they needed to see eight tigers before they would worry, they would likely be dead. However, if you flip a coin three times and get heads each time, you wouldn't assume it's a special coin, would you? It's important to have sufficient facts and maintain a scientific approach. The answer is eight because that's the point at which it's less than 1% likely to have happened by chance. If an event is very unlikely and it happens several times in a row, then you should be suspicious. However, we're only talking about figures slightly above the average, which is only 50% unlikely, and that's where you use this figure of eight. If times, complaints, or costs exceed the average eight times in a row, the probability of this happening by chance is one in 256, or about half a percent. Therefore, we can be slightly over 99% certain that something has changed, and there is less than a 1% chance that it is due to pure bad luck.. 

Scene 7

[Audio] We have already established that if an event has a 50/50 chance of happening, such as something being above average, you need at least eight occurrences before you can be certain that something has changed. However, if the event is unlikely, you need fewer occurrences to detect a change. In fact, the question can be reversed: how unlikely does something have to be before a single occurrence warrants an investigation? What is the warning level at which action and investigation should be taken? Credit card companies use this calculation to prevent fraud. It may be inconvenient, but it is necessary. How many complaints would prompt an investigation? How many complaints would prompt an investigation? What is the threshold for going over budget? What is the threshold for accidents or sick days in this area of your business? At what point would you investigate a drop in sales for just one month? It is important to determine trigger points for any measured data. This information is crucial for effective measurement and analysis. Using the formula of the mean plus or minus 3.14 times the MMR (median moving range) can help establish these trigger points. The following is a column of numbers that represents the number of accidents that occur each month at a railway station. Please note that this does not include train crashes, but rather minor incidents such as slipping on a patch of spilled coffee. Nowadays, every little thing is recorded. If I obtain this data from one hundred different railway stations and need to monitor them to determine whether I should investigate a particular station, I will perform the following calculation: Do not worry, I will explain. First, I will calculate the mean, which is 7.9. Calculating the average is simple: add up all the numbers and divide by the total count. To find the median moving range (MMR), arrange the absolute differences between each consecutive number in ascending order and select the middle number as the MMR. In this case, the MMR is four. Improving this text was straightforward. You can automate the process using a spreadsheet or computer. The median moving range (MMR) is the middle number and represents the typical amount that the number changes from one day to the next. The content of the improved text is as close as possible to the source text, and no new aspects have been added. A small MMR means that even a slight change in the number of accidents will have a significant impact, while a large MMR requires a substantial change in the number of accidents before any concern is warranted. Simple vocabulary has been used to ensure that the text is accessible to a broad, general audience, including those with cognitive disabilities, low reading literacy, or who are encountering an unknown topic or language. The content of the improved text is as close as possible to the source text, and no new aspects have been added. In this example, the MMR is four. Therefore, the action point is approximately 12.5 above the mean, as indicated by the formula. To work it out, find the gaps, identify the middle-sized one (MMR), multiply it by 3.14, and add it to the mean of 7.9 plus 12 1/2, which is about 20 1/2. If I hear of 21 accidents. 

Scene 8

Scene 9

[Audio] In every process, there are three factors to consider: quality, cost, and time. Quality refers to the standard of the product or service being produced, while cost refers to the expenses incurred during the process. Time, on the other hand, is not simply the number of hours spent on the work, as this falls under cost. Rather, it refers to the leadtime - the amount of time customers have to wait after placing an order. Time can also refer to on-time delivery, although this is usually affected by variability, which is more closely related to quality. If tasks vary each time they are performed, they may need to be repeated, which can affect on-time delivery. Additionally, producing things quickly can also be costly, which I will explain shortly. Therefore, the root of your processes is quality, cost, and time. Unfortunately, it is impossible to have all three simultaneously. Reducing quality can lower costs, but higher quality usually takes longer to produce. Many organizations fail to consider the implications of their desire to provide high-quality products or services at a low cost and with quick turnaround times. Sales representatives or advertising campaigns may promise such outcomes to customers, but the plan to achieve them often involves putting excessive pressure on employees or assigning a single person to handle multiple tasks. This approach is not sustainable and can lead to negative consequences. Sales representatives or advertising campaigns may promise such outcomes to customers, but the plan to achieve them often involves putting excessive pressure on employees or assigning a single person to handle multiple tasks. The text appears to be error-free and already meets the desired characteristics. Therefore, no changes have been made. There is a risk that the mixture may not be optimal. The strongest party will prevail, or you may end up with a mediocre combination of quality, cost, and time. Depending on the market, it may be more appropriate to aim for maximum quality or maximum speed of delivery, and make a conscious, realistic decision to be less good at the other two factors. In most markets, there has been a shift from cost to quality and then to time. Until the 1950s, the main focus was on being the cheapest. However, quality became an issue and Japanese cars transformed from being cheap to high quality. As a result, low-quality producers mostly went out of business. Nowadays, the focus is mainly on time. Nevertheless, the Chinese have shifted the emphasis back to cost due to their lower wage costs and economies of scale, which allows them to produce things incredibly cheaply. Cost is currently a focus due to the ease of shopping around online. However, this may only be a short-term trend. Research suggests that high-quality producers, such as German cars or Apple laptops, gradually gain market share and make more profit per unit, resulting in the highest overall profit. The next section will explore the relationship between quality, cost, and profit. We will revisit lead times later. Before that, consider your organization or market. Is it cost-sensitive? Should you prioritize low costs and prices, high quality, or rapid delivery times? What are your competitors doing? Do you need to change your overall strategy?. 

Scene 10

[Audio] Does it cost more to make a better product or provide better service? At first sight, you might think so. However, there are two parts to the cost of quality. The first is prevention cost, which increases as quality improves. Prevention cost includes ensuring that things are done well, such as training, quality control systems, good management, excellent equipment and materials, and employing skilled personnel. Making everything perfect can be very expensive. In fact, the cost increases exponentially as you approach perfection. On the other hand, the cost of poor quality increases as the quality decreases. The more substandard the quality, the more you have to pay for scrapping, reworking, handling customer complaints, losing contracts, and even facing legal action. Using cheap materials, equipment, and labour may seem like a cost-saving measure, but it can actually end up costing more in the long run. By combining the two graphs, we can see the total cost, which includes both prevention and failure costs. By combining the two graphs, we can see the total cost, which includes both prevention and failure costs. By combining the two graphs, we can see the total cost, which includes both prevention and failure costs. As shown on the right, failure costs have decreased to almost zero, resulting in a lower total cost. Similarly, on the left, we were not doing any prevention, and the only cost was the high failure cost. In the middle, where the two graphs intersect, our failure cost equals our prevention cost, and the total cost is double each of these. Therefore, the total cost has a U-shape, with the minimum cost point in the middle. This is the most cost-effective point. I will explain shortly why being here is not desirable. However, it is worth noting that many organisations fail to recognise this as the most cost-effective point because they do not measure their failure costs. They are acutely aware of their prevention costs, such as training, quality control, and the use of high-quality materials and machinery. There is always pressure to reduce these costs. Budgets for these costs are monitored and cut slightly every year, which creates pressure to reduce quality to save money. However, they often overlook the invisible costs of failure. There is no budget for rework or handling complaints, and the cost of lost customers or wasted staff time is not measured separately. Therefore, it is essential to aim for the point of minimum cost, rather than the point of maximum profit. Moving from five to about seven reduces costs and increases profit. The point of maximum profit is to the right of the middle on a sales volume or price graph. There are four important learning points to take away from this. Choosing low quality is a false economy. The point of minimum costs is not the best place to be, as it does not yield maximum profit. The point of maximum profit is at about seven. To get there, it is important to consider how to achieve the desired level of quality without exceeding the cost-benefit ratio. It is not worth trying to reach a quality level of 10. Finally, the question remains: how can we create this graph in reality? How can we determine our position on the theory? The answer lies in listing and estimating prevention and failure costs. Investing in better quality, whether it's people, equipment, materials, or methods, may prove to be cost-effective. If your failure costs are greater than your. 

Scene 11

[Audio] In any company, there is inevitably some waste. For instance, during meetings, how much time is actually productive? Similarly, how much of your working day is productive? Time is crucial because your biggest expense is likely to be your employees, which is essentially their time. Additionally, what proportion of materials are discarded, and how much does that cost you annually? Expensive resources such as machinery, laboratories, and even office space - what is their utilization? Do they use everything all the time, without wasting space on piles of unused items? According to a Japanese theory, there are seven types of waste, but I prefer to remember them using the word 'downtime', which includes an eighth type. The 'D' in 'downtime' stands for defects, which can cost you money even if you discover them immediately. The letter O in downtime refers to overproduction, which occurs when more products are made than necessary. This may happen when items are produced in large batches or for stock, resulting in the need for storage and inspection, and tying up capital. The letter W in downtime stands for waiting, which refers to the time that a product spends waiting, rather than the time that people spend waiting. This is a half-finished product that may be waiting for a machine to be set up or for other parts in the batch to be worked on. It may also be waiting for a broken machine to be repaired. This waiting time is what distinguishes the short time it takes to make the product from the full lead time that the customer has to wait for it. Remember, this applies not only to manufacturing but also to services. For example, waiting for the keys to your rental car is a service, and the keys are the product you are waiting for. Consider the time spent waiting for them. The 'N' in downtime refers to unused employee talent, which is the eighth waste. Every talented employee is a valuable resource that should not be wasted. Many organizations fail to tap into the full potential of their employees, resulting in a significant loss. It is common to find individuals who are disengaged from their work because their employer has failed to motivate them. However, these same individuals may be highly engaged in their personal lives, coaching sports teams or engaging in other impressive activities. They have spare energy and potential that work isn't using. Downtime includes transportation, which is the movement of components, paper, or equipment around the site. This can add up if you follow a job from start to final dispatch. Inventory refers to raw materials, work in progress, or finished materials ready for shipment. It is important to keep inventory levels low to avoid tying up money and incurring storage costs. It is important to minimize motion to ensure efficient operations. Downtime has two components: inventory and motion. Motion, on the other hand, refers to wear and tear on machines and people, and the risk of accidents and damage. Finally, the 'E' in downtime represents excessive processing. This includes over-design, performing operations on parts of the product or service that the customer won't see or care about. Ensure the text is grammatically correct, uses simple vocabulary, and is accessible to a broad audience. It is important to avoid adding new content or changing the meaning of the original text. To minimize these issues, it's important to review every stage of your processes with these factors in mind. Remember to keep sentences. 

Scene 12

[Audio] Quality, cost, and time are all important factors to consider. However, lead time is often overlooked. It is assumed that working harder will make things quicker, but there is actually a lot of interesting science and theories behind lead time. In this text, we will explore these concepts further. It is assumed that working harder will make things quicker, but there is actually a lot of interesting science and theories behind lead time. The easiest way to ensure quick product delivery is to keep them in stock. It may seem like the time it takes to make the product is irrelevant if you have stock available, but in reality, stock is often a poor choice. It requires space, can expire or become damaged, needs management and checking, ties up capital, and can lead to decreased manufacturing efficiency. Keeping stock can limit flexibility and hinder the ability to meet individual customer needs. Accurately forecasting future demand can be challenging, and not having an item in stock can be problematic. Therefore, there are hidden costs associated with keeping stock that are often overlooked. Additionally, services cannot be stored in stock. To ensure timely production, items ranging from hot pizzas to heart transplants cannot be kept on the shelf. Therefore, it is important to consider whether items currently kept in stock could be made to order instead. Although this may increase costs, it could ultimately save money by reducing the need for excess inventory. Whether you are producing for stock or customers, the question remains: how can you reduce lead times? Why does it take so long to make things? I used to work in a helicopter factory. It took three years to make a gear wheel for a helicopter gearbox. There were only two weeks of actual metal cutting.The rest of the time was spent transporting stuff around the factory and waiting to start the batch. Often, there were 10 other batches also waiting for that machine. Then, waiting for the rest of your big batch. Some batches were 500 parts before finally your part gets made. If you follow the piece of metal, you will find that it undergoes four main activities. Firstly, it transfers from one part of the process to the next. Secondly, it waits in a queue to be operated on. Thirdly, it waits while the rest of the batch is operated on. Finally, the only value-added activity in all of this is when it is operated on itself. The same principles apply to an office environment. Therefore, the lead time could be the time from receiving a planning application to approving or rejecting the new building. It could also refer to the time it takes to get from the airport to the plane, or to service your car, or to receive your pizza in a restaurant. Additionally, it could refer to the time it takes to create a training course like this one. All of these are lead times. Queues now exist as piles of paper on desks, in email inboxes, and as waiting lists. A six-month wait for a doctor is essentially a six-month queue of people. To expedite processes, there are three methods: firstly, improve the layout to enable faster movement of items between machines, people, or departments; secondly, use active voice to make instructions clearer; and thirdly, use verb phrases instead of noun phrases. There may not be significant savings, but it is worth considering, particularly for large-scale projects such as car manufacturing.. 

Scene 13

[Audio] To streamline any process, the first step is to consider reducing the number of steps. This is part of Business Process Reengineering (BPR), which involves examining the process and creating a flow diagram. Then, ask four questions: Can any steps be deleted? Your process may have more steps than you realize. Forms are filled out, data is entered into computers, work is assigned to people, and supervisors check for quality. Consider whether the supervisor's check is necessary, if the manager can input their own data, or if the order form needs to be printed. Another aspect of BPR is automating steps through the use of computers, which can be just as effective as eliminating them. Drawing out the entire process and consulting technical experts can help identify opportunities for automation. Part three of BPR involves identifying steps that can be combined and assigning them to a single person. For example, the person who receives a phone call can also be responsible for resolving the issue. Can a designer also purchase the parts they have designed? Can a machine operator also pack their work? It is surprising that some lawyers still dictate emails to their secretaries. Consider if there are any opportunities to combine stages in your systems, which could eliminate a transfer and a queue. Although person two may be faster at completing your job, it is still likely more efficient for person one to complete all the tasks rather than waiting for person two to finish. Additionally, the fourth aspect of BPR involves identifying steps that can be completed in parallel rather than sequentially, which can reduce the number of steps and speed up the process. It can save you money to eliminate unnecessary steps or automate them. Some manufacturing companies have taken this to the extreme with production cells. Volvo conducted an experiment where a team of people built an entire car from start to finish, instead of using a conventional flow line where the car is passed along to different stations for each step. Although the team was not as fast as specialists in assembling the car, they saved time on transfer and waiting. The car was worked on continuously, and the team was dedicated to ensuring quality improvement. So, in the idea of a production cell, a small group of people, often sitting or standing in a U shape, and by the way, they're on the inside the U of tables, so they can swap around and help each other, they process a job from the start to the finish. Additionally, using active voice and simple sentence structures can make the writing clearer and more accessible to a broad audience. It is important to avoid adding new content or changing the meaning of the original text. Setting up teams or cells can increase productivity and efficiency, especially when assembling similar items. Consider creating cells for family groups of items that need to be assembled. Although it may not always be efficient for each individual, sometimes we have to sit through agenda items that do not affect us or require only brief input. However, this approach is very efficient for the task at hand. Instead of being passed from one desk to another for weeks, it is dealt with by everyone in a few minutes, and the job is done. Therefore, there is a trade-off between individual efficiency and processing jobs within a short lead time. Drawing out and reviewing. 

Scene 14

[Audio] In the office dining room, you receive your soup in 20 seconds. Next, you move to the meat counter where a nice piece is cut for you in just 3 seconds. After that, you get your vegetables in 20 seconds. Finally, one server provides you with both your dessert and drink in 15 and 10 seconds respectively. The payment process takes a minute. Take a moment to review the diagram. Can you determine how many lunches this setup can serve per hour?This question highlights a crucial and straightforward concept. Avoid getting bogged down with a calculator and numerous calculations. What is the bottleneck of the system? The cashier is the slowest part and can only process 60 people per hour. The number of people you can serve depends on what others are doing. What is causing the delay? Identifying the bottleneck in your work is crucial. As the manager, how can you enhance the current setup without investing in additional equipment or personnel? Please note that self-service is not permitted, and portion control must be maintained. Remember, if you concentrate on the bottleneck, the solution becomes apparent. There is no value in performing tasks faster than the bottleneck;it is a waste of capacity. To improve efficiency, remove anything that is faster than the bottleneck. This will allow for the redeployment of two people, resulting in a team of three. With this reduced team, they can still serve all the food within the rate of one person per minute, which is the speed at which they are currently processing orders at the cashier.If you have more demand than one person per minute, it is better to serve the extra customers by speeding up the bottleneck. Let's suppose you have two cashiers. Now you can serve a person every 30 seconds, which is 120 people per hour. By adding just one extra cashier, you can sell an additional 60 lunches, which at a cost of $10 per meal, would result in an extra $600. This demonstrates the value of working on the bottleneck, as it frees up the entire system. Consider how this concept could be applied. Finally, returning to the original setup, how long will it take to receive your lunch? This question may seem easy, but it is actually quite challenging. Please give it a try and see if you can come up with the correct answer. You will feel accomplished if you succeed.. 

Scene 15

[Audio] Understanding the bottleneck idea and queues is crucial as it applies to any operation, including yours. There are three perspectives to consider when determining how long it takes to get lunch. The first is from the viewpoint of the first person in the dining room. They go through all the stages, and it takes a total of two minutes and 35 seconds to get served. However, I do not have any concern for them as they are a unique case. What about the rest of the people? The situation is different for them. Therefore, the second approach to consider is to envision a conveyor belt where everyone moves once per minute.This is the scenario when all the trays slide on rails. Getting your soup takes 20 seconds, but you have to wait a whole minute before moving on to the meat because all the trays are waiting for the cashier to serve their person. This means it takes a total of five minutes to get your lunch. Another way to think about this problem is to consider what would happen if there were islands. Imagine visiting the soup island, enjoying a bowl of soup, then moving on to the meat island, and finally reaching the cashier's island. This analogy highlights the flow of jobs between different offices or departments. It takes only 20 seconds to get a bowl of soup, allowing three people to be served per minute. They will go to the meat island where only two people are served per minute due to the 30-second time it takes to prepare the meat. As a result, the queue at the meat island increases by one person per minute. Two people arrive at the cashier per minute, but only one person is served per minute, causing the queue at the cashier to also increase. After 10 minutes, there are 10 people waiting at the meat island, resulting in a 5-minute wait to receive their meat. Additionally, there are 10 people waiting at the cashier, resulting in another 10-minute wait. In total, there is a 15-minute wait, not including serving times. This situation is similar to jobs piling up in the accounts department or people queuing for airport security, where queues can become very long. Islands are useful for those who want to skip certain food items, but they can cause uncontrolled queues and delays. Meanwhile, the conveyor belt system puts pressure on the slowest point to speed up, which in this case is the cashier. Meanwhile, the conveyor belt system puts pressure on the slowest point to speed up, which in this case is the cashier. In a work setting, management or those being held up could assist. There are three main conclusions from this example. Firstly, it is easy to determine the system's delivery capacity by identifying the bottleneck, which in this case is the cashier. And finally, the text should flow logically and any processes should be described in order. Secondly, the bottleneck restricts the throughput. Either plan everything to go at that speed or speed up the bottleneck. Investing a little more in the bottleneck can yield significant gains for the entire system. In our case, adding a second cashier cost only an extra $10 per hour, but we served an additional 60 meals, resulting in a $600 profit. Therefore, it is essential to take care of the bottleneck. Additionally, it is crucial to consider the time spent in. 

Scene 16

[Audio] In the previous section's lunch line example, we observed how the cashier was the bottleneck. Identifying and enhancing the cashier's throughput made all the difference. The concept of bottlenecks originated in the 1980s with the publication of Eli Goldratt's book 'The Goal.' Goldratt's book had a significant impact on me, as well as many others. His first point is that you cannot produce more than the bottleneck can handle. Knowing the bottleneck allows you to determine the maximum production capacity, regardless of other factors in the factory or office. Increasing the bottleneck's capacity will increase the overall system capacity. Therefore, investing in the bottleneck is worthwhile, while investing in non-bottleneck areas is a waste. The bottleneck concept leads to many subtle ideas, and here are three. To avoid work or people piling up in your system, it is important not to pushing more than the bottleneck can handle. Use this knowledge to schedule the input of work or materials, as well as what will come out of the system. If the bottleneck can only process 10 per day, then the whole system will only produce 10 per day. There cannot be more than 10 items produced, nor can there be less. If there is a temporary delay downstream, you can catch up the following day and return to the average of 10. Another concept that arises from bottlenecks is that time lost at non-bottlenecks is not truly lost because it can be made up as long as it does not affect the bottleneck itself. Ideally, there should be a buffer of work in front of the bottleneck to ensure that it never runs out of work to do. Therefore, upstream issues are irrelevant as the process continues to work on its buffer, and downstream issues are also irrelevant. Therefore, management efforts should focus on ensuring that the bottleneck never fails, always has sufficient work to do, and is never supplied with defective parts or materials. The third important concept from bottleneck theory is to avoid having operations that require significantly more resources than the bottleneck. This is wasteful because the output will never get through the bottleneck. However, do not reduce these operations to the same level as the bottleneck, as this will create two bottlenecks. If the bottleneck is eliminated, the next lowest production area must be identified .Every system has a bottleneck, so what is yours? How do you find your bottleneck? There are three ways to approach this. Firstly, calculate the required hours for each stage to handle the total throughput and determine if there is enough capacity at each stage. For instance, if it takes one minute to attach the wheels to the car and 600 cars are produced daily, do we have 600 minutes, which is equivalent to 10 hours, of available personnel at the wheel attachment stage? Secondly, to identify the bottleneck, a much simpler method is to locate where the queues are. If there is a queue for the photocopier, it may be the bottleneck for your office. Similarly, if the booking-in clerk has a pile of paper on their desk, they may be your bottleneck. It is unacceptable if one of your cheapest resources is strangling the output of the whole operation. To identify a bottleneck in a system with multiple products and paths, examine which items tend to be delayed and locate where they are being held up. This is likely to. 

Scene 17

[Audio] Queues are the primary cause of long lead times, and they build up in front of the bottleneck. When you have enough capacity, the queue will continue to grow if the number of arrivals exceeds the number of people served. For example, if you can serve 10 people in an hour, and 12 people arrive per hour, the queue will grow by an average of two people per hour. After 20 hours, the queue will have 40 people, which is undesirable. If you can serve 10 people in an hour and only eight people are arriving, it may seem like there won't be any issues. However, the problem arises because the eight people won't arrive evenly spaced. This results in a queue forming when several people arrive at once, followed by a gap where no one is being served. As a result, your capacity to serve 10 people per hour is reduced to maybe only seven. The first problem is fluctuating arrival rates, which also leads to fluctuating service rates. This means that sometimes it takes only four or five minutes to serve a customer, while at other times it takes seven or eight minutes. The latter group of customers reduces overall output. Additionally, when a customer only needs five minutes of service, there will be a gap in which capacity is wasted and employees are idle. When planning the necessary capacity to provide good customer service, it is not sufficient to simply calculate the average number per hour. It is important to consider the amount of spare capacity required to handle fluctuations in arrival and service rates. Fortunately, there is a formula available to determine the necessary extra capacity. The formula states that the queue length is equal to the utilization divided by one minus the utilization. If percentages are used to represent the utilization, such as saying 75% of the time they are working instead of saying they are three-quarters utilized, the formula for predicting the queue length will be the utilization divided by 100 minus the utilization. Keep in mind that this is only mathematics, so there is no need to panic .This formula is easy to use and applies to situations where work arrives at a random rate and you need to serve the customer quickly. For example, an IT help desk customer service line where it takes seven and a half minutes to deal with the call. On average, you should be able to serve eight people per hour .If you have 64 calls coming in per hour, you would expect to need eight people to handle the demand. However, there will be times when there is no demand, resulting in wasted capacity and fewer than eight people being served per hour. On average, 64 items arrive per hour, causing the queue to grow. To prevent this, spare capacity is necessary. But how much? If we increase the number of workers from eight to ten, will that be enough? Let's use the formula. First, we calculate our utilization. With ten people doing the work of eight, our utilization is only 80%.According to the formula, the queue length will be U divided by 100 minus U. In this case, it is 80 divided by 100 minus 80, which equals 20.Therefore, the queue will have four people. As each person takes about seven minutes to serve, the waiting time will be approximately 28 minutes. This may be considered too long. Suppose we still receive 64 calls. 

Scene 18

[Audio] The basic idea is to avoid producing work that the next person in the chain cannot handle. Buying food from a supermarket or ordering books from the internet are both examples of pull systems because you purchase items at your own pace. The e-mail inbox, post, and most factories operate on a fixed delivery rate, regardless of demand or capacity. The pull system aims to prevent upstream work unless the downstream can handle it. There are two simple ways to accomplish this. The first is to implement an ordering system, similar to that of a supermarket. The second involves sending a message from one person to another, requesting either more production or delivery. This message may be in the form of a Kanban card, which includes the part number and required quantity. The first person delivers the parts with the card. The second person can then use the card to reorder when they need more parts. Perhaps the card is affixed to the front of the transport box to indicate the parts, or it is placed at the bottom of the box for easy reordering once all the parts have been used up. The crucial point is that the system cannot contain more items than the number specified on the card or the number of boxes in the system. Person one cannot commence any work until the card has been received. The system is simple and effective, even when dealing with a large number of contributors. To maintain a manageable workload, a storage area can be implemented between the two parties. Person two retrieves work from the storage area, and person one can only produce more work when there is space in the buffer. This ensures that if person two stops for any reason, person one must also stop. Although this may not sound ideal, it is an effective way to manage workflow. Remember, there is no benefit in person one producing more than person two requires. Instead, person one should assist person two. This is an interesting and subtle additional advantage of maintaining low stock levels with a pull system.As there is only a small buffer between them, the two individuals are closely connected, and they must work more collaboratively. Likewise, if person one encounters an issue, person two will run out of materials and will need to check on person one's well-being and offer assistance. Improved teamwork and capacity balancing are achieved through a pull system. For instance, if someone has more than 20 e-mails in their inbox, you cannot e-mail them. In this case, you can either assist them with their workload or contact a colleague with available inbox space. This approach can be an effective way to work. Why contact someone who is already busy when you have other options? Once the pull system is functioning and buffers are being replenished smoothly, buffer size can gradually be reduced. To summarize, why should person one continue producing when person two cannot keep up with the supply rate? Firstly, it can result in a backlog of work and longer lead times as items have to wait for the next person in line. It is important to have a clear process in place to address these issues. Secondly, it can mask underlying issues such as inconsistent quality or high defect rates, as the next person in line can simply sort through and select the best items. There are two main reasons why having multiple people involved in a. 

Scene 19

[Audio] Earlier, we learned that smaller batches result in less stock or work in progress, along with other benefits. However, the reason for batch production is to reduce setup costs. If it takes an hour to set up the machine and then one part can be produced every second, how many parts can be produced in a batch?Take a moment to consider before I reveal the answer. It takes an hour to set up the machine, but once set up, it can produce one item per second. If you only need one item per month, then setting up the machine for an hour would produce 12 items, which is much more efficient than producing 3,600 items. The production rate is affected by the usage rate, and storage cost is another factor to consider. For example, if we're talking about sofas, it wouldn't be practical to have 3,000 of them sitting around. The EBQ, or economic batch quantity, is a formula that was first developed 100 years ago. It suggests that the optimal batch size is the square root of two times the annual usage, multiplied by the setup cost, and divided by the annual storage cost per unit. However, many people today consider this formula irrelevant. Don't spend too much time learning it. The current belief is not that setting up takes a long time, requiring a large batch, but rather questioning why the setup time is so lengthy. When the idea of reducing setup time, possibly to almost nothing, is considered, it alters everything. I recently saw a group of people change a car engine in just 58 seconds on TV. They drove the car into the studio, uncoupled the old engine, lifted it out, replaced it with a new one, and drove the car out of the studio in under a minute. This impressive feat raises the question: why does it take an entire hour to change jobs on a milling machine or switch out tools on an injection moder? Filming people during the process reveals that they struggle to locate tools, often need to retrieve additional equipment, encounter breakages, struggle with lifting, and require a pallet truck. Additionally, it may take several attempts to achieve the correct adjustment. Streamlining these processes can result in significant time savings during batch changes. You can make smaller batches efficiently, saving space and lead-time throughout your process. The same principle applies to purchasing. The economic order quantity (EOQ) formula can tell you how many items to buy at once. As previously mentioned, the cost of usage, ordering, and storage are taken into account. However, the primary concern should be why a system cannot be implemented to allow for smaller quantities to be ordered and delivered. Although it may be more expensive to produce, the savings on shipping and other costs make it a viable option in certain cases. Even in office settings, work is often batched due to the cost of changeover. Does the person who processes invoices store up a week's worth and then do them all on Fridays? Could they process them daily to ensure suppliers receive payment promptly? Purchasing is often centralized to buy items at a lower cost. However, this practice can result in hidden costs such as waiting and administration. Therefore, reducing set-up time not only saves time but also allows for smaller batches, reducing lead time and stock holding costs. Formulas such as EBQ and. 

Scene 20

[Audio] Quality isn't just the job of the quality department. Ideally, the person at the top of the company would be obsessed with quality, seeing it as a way to beat the competition, get happier customers and reduce costs, instead of just a pain that means that some things don't get shipped because they fail inspection. And as well as support from the top, quality should involve everyone in the company. Often, production people see quality just as inspection and audits to be passed every now and then. Well, actually, these production people are the most valuable process improvement asset that you've got if you could motivate them to care about quality. Back in the 1960s, Professor Ishikawa, in Japan, had the idea of a quality circle, where management would actually ask the workers for help and ideas. This approach was considered radical when it was first introduced, and it may still be seen as such in some companies. The concept involves assembling a team that includes individuals from all levels and disciplines to work on a specific problem and find a solution. This approach recognises that ordinary people, regardless of their education or position in the hierarchy, often have the best ideas because they have a deep understanding of what really goes on. They are also motivated by being asked to contribute to a team effort that addresses an important problem. Quality circles were successful in the 1960s, but have since been forgotten. Continuous improvement, or Kaizen, has replaced them. The idea is that if everyone improves their part of the process by a small percentage, the entire company will move towards perfection. Toyota implements 5,000 improvement suggestions per week. They implement all 5,000 suggestions they receive. Their workforce is always brainstorming ways to improve things. While most Western companies have suggestion schemes, they typically only receive about 20 ideas a year, half of which are anonymous and rude suggestions about the boss. This is a real problem for continuous improvement in Western companies, as it can lead to boredom with the process. If you are struggling to maintain regular process improvement meetings, consider setting up a quality circle to investigate and fix your biggest problem. Provide them with the necessary time, money, and empowerment to achieve their goals. By doing so, you will be surprised at how much they can accomplish and how much they will enjoy it. Therefore, think about your most significant quality issue and who you could include on the quality circle team.. 

Scene 21

[Audio] Two final thoughts. Firstly, a culture of great quality and service, which likely results in shorter waiting times, must come from the top, even if it requires increased costs. However, it will be worth it. Sometimes, you have to spend money before you can save it or sell more and bring in more revenue.This act of faith requires senior managers who understand queue theory and the true cost of poor quality. If needed, you may be able to persuade them to consider this cost. My second point is tonsure the accuracy of data and personnel before addressing system issues. Reduce work in progress and focus on bottlenecks to implement Lean principles and improve quality.. 

Scene 22

Process Improvement Foundations

Scene 0

Process improvement involves measuring the core of your company's operations and finding ways to enhance it. This could mean achieving better quality
 doing it more cheaply
 or doing it faster. Improving your processes can increase profitability and is often necessary for survival in a competitive market. Although every company and business is unique
 there are techniques that can be used to investigate process improvement. This course
 led by Chris Croft
 will cover measuring the current process
 evaluating its effectiveness
 considering the trade-offs between cost
 time
 and quality
 identifying bottlenecks
 and exploring other relevant theories. By the end
 you will have a better understanding of how systems work. You will also have practical tools to improve your own processes. Let us begin.

avatar

To improve a process
 first determine if it needs improvement. When intervening
 it is important to assess whether changes are due to your actions or external factors. It is crucial to measure the effectiveness of any changes made. Even without intervention
 there is a 50-50 chance of improvement from one year to the next. Therefore
 it is necessary to monitor processes. It has been said that if something cannot be measured
 it cannot be managed.     However
 is it really impossible to manage something without quantifying it? Can everything be measured
 including quality
 service
 design
 or culture? While some may argue no
 as an engineer
 I believe that everything can be measured. Can everything be measured
 including quality
 service
 design
 or culture? I believe that measuring is crucial to avoid assuming everything is okay when it might not be. Without measuring
 problems may go unnoticed until they become visible to everyone. I believe that measuring is crucial to avoid assuming everything is okay when it might not be. Therefore
 on this course
 I suggest identifying key metrics and monitoring them to optimize performance. Optimize does not mean maximize or minimize. It is often best to aim for 80% rather than 100%
 as we will see. 
Scientific management is the starting point. Determine your desired outcome and measure progress towards it. Hearing a company announce a reorganization can be disheartening
 as it often leads to years of costly chaos
 with valuable employees departing and less competent ones remaining in hopes of receiving redundancy pay. To improve a process
 it is important to measure everything and identify problem areas. Target those areas and measure progress to ensure improvement. Avoid making changes that do not address the identified problems. What are black holes? How can we measure these seemingly immeasurable processes in your field of work? If measurement were possible
 what methods would you use?

Measuring everything can be overwhelming and lead to confusion about which numbers to focus on. Self-contradictory measurements can also arise when trying to increase quality and reduce costs simultaneously. It can be challenging to determine which metric to prioritize.Ensure that cost reduction does not compromise quality or other important factors. Use simple vocabulary and sentence structures to make the text accessible to a broad audience.Always consider context when interpreting numerical data.  For example
 an increase in turnover does not necessarily mean an increase in profit
 as prices may have been cut to boost sales. Avoid adding any new content to the text. Similarly
 an increase in the number of employees may not always be a positive sign and depends on the corresponding increase in turnover. Complaints may have increased because you are improving your customer service or undertaking ambitious projects.     Numbers can be deceiving. For example
 if you hear that your local school is 5% worse than the one in the next neighbourhood
 you may feel unhappy. It is important to avoid relying on a single metric. However
 it is important to consider what 'worse' means and whether a 5% difference is significant. Was it just a blip when it's usually better than other schools? And maybe your school is worse at one thing
 but they're better at other things that are more important. How do we know that the school down the road isn't just better at playing the game so they get a better score at the expense of areas that aren't measured? A recent example of this in the UK was the government trying to measure doctors. And they published a league table of scores
 but it was discredited and then discontinued because some quite important factors weren't included in the scoring. And also the weighting of the factors was controversial. And it was hard to take into account the effect of living in richer and poorer areas. And most importantly
 the doctors who were willing to take on the most difficult patients got lower scores
 even though they were probably better. So should we give up and ignore measurement? Well
 that would be a shame. Measurements are crucial
 but they should not be used to judge people. Instead
 they should be used as a starting point to ask questions. For example
 why does one doctor have a lower score than another? Why does one school achieve better results? By understanding the causes of the scores
 we can determine their significance. To do this
 it is important to start by obtaining a ratio. To interpret statistics like infections per operation
 days absent per pupil
 accidents per passenger mile
 and errors per report
 it is essential to compare the ratios with similar people
 departments
 companies
 or with oneself at the same time last year. Comparing ratios is the only valid way to understand such statistics. Using the example of 20 infections
 it is not enough to understand the severity of the situation. However
 20 infections per hundred operations is a better indicator
 but still lacks context. When everyone else has only 10 infections per hundred
 you're making progress. Therefore
 ratios must be used to make comparisons.

Consider the source and whether there may be any bias. 1) Is there potential for bias? Additionally
 it is important to look at the whole picture and determine if there is a causal relationship between variables. To evaluate information
 use these five tests: When reviewing data
 it is important to question its reliability. 2) Is it a number or ratio? 3) Is it significant?  Could this have happened by chance? Does the number vary significantly between time periods
 or is this just a chance occurrence? What other factors are they not measuring? Lastly
 can we be certain that one thing caused the other?   Establishing causation can be challenging
 as numbers are often correlated
 meaning that if one increases
 the other may also increase. However
 they are not directly causing each other; rather
 there is a shared underlying cause. For instance
 children from educated households tend to read more and perform better in school. However
 it is not necessarily the act of reading that leads to their success. Other factors
 such as their attitude in the classroom or the homework they complete at home
 may also play a role. It is crucial to distinguish between these factors because if a correlation is found between reading and academic achievement
 one might assume that reading is the sole cause. It is unclear whether every child should read more.   Proving causation between reading and improvement is difficult and requires a controlled experiment. For example
 comparing middle-class children who read to those who do not
 or comparing a group of children from a deprived area who read more to an identical group who do not. Simply finding an existing correlation is lazy and does not prove anything. Similarly
 discovering what leads to increased sales at work can be challenging. If you reduce the price and sell more
 it could have been caused by something else. Did anything else change at the same time? Therefore
 as an exercise
 consider what you are currently measuring. Are you measuring only one thing out of context? Are you measuring too many things? This can make it difficult to know which ones to focus on. Are you measuring ratios rather than straight numbers? And are you comparing them with similar people or time periods? And are you sure about your causations?

Statistical process control (S-P-C--) is a simple yet useful concept that is often taught with a lot of maths.   The principles behind S-P-C are straightforward and easy to understand. The idea of process control is to inspect things during the production or service delivery process
 rather than after it's completed. For example
 checking how well a room is being cleaned or how well a meal is being cooked during the process
 rather than after it's finished. Learning from inspections is helpful
 but it is not as effective as having control over the processes. It is better to ensure that the oven is at the correct temperature and the turkey is cooked for the appropriate amount of time
 rather than discovering that it is burnt. Therefore
 we should focus on getting the temperature and timing right every time. We could try using a hotter oven for a shorter time or a cooler oven for a longer time to determine the optimal way to cook a turkey. Once we find the perfect method
 we should stick to it consistently. This is an example of process control
 which is simple for cooking a turkey but more challenging for tasks such as cleaning
 writing software
 or creating a training video. This is where statistical analysis becomes necessary. To ensure success
 it is essential to identify the critical components of the process and ensure they are executed correctly. There are two main reasons for failure: drifting of the average or too much variation. If we follow the recipe but our turkey is burned
 it means that either the oven is out of calibration or the temperature is fluctuating in the oven. Which one is the cause? To determine whether it is drift or variation
 take four random readings and observe their dispersion. If the readings are all very close
 it indicates that the oven temperature is calibrated. However
 if the readings are scattered
 it suggests that the temperature control is failing. Additionally
 this sample of four provides an average measurement
 without being influenced by occasional anomalies. By taking samples every hour or every day
 we can monitor whether the average is drifting. The sample of four enables us to monitor both drift and variation. We can then plot these on graphs to visualise the process over time.This approach not only explains why we failed and how to fix it but also provides advance warning of potential issues. For instance
 if our oven is starting to drift out of calibration
 the S-P-C graphs will alert us well before we can detect any issues with the turkey. We can avoid the embarrassment of a disastrous Christmas dinner or sending incorrect items to customers by implementing proper measures. Consider tracking the time taken to serve customers and the number of typos in reports. It is important to be able to measure if things are getting worse before a complaint is made.   To apply statistics and draw graphs
 the measurement must be quantifiable. The pass/fail system is not sufficient as it does not provide a clear understanding of what is happening in the process. The question remains: how can S-P-C be applied to your part of the business? What are your key processes? What are the indicators within those processes that you could measure and track on graphs over time?These indicators could include cost
 waste
 hours taken
 overall delivery lead time
 percentage delivered on time
 quality
 accuracy
 or customer satisfaction.   Using Statistical Process Control (S-P-C--) is highly recommended as it is one

How can we determine if a machine
 process
 or department is truly out of tolerance
 indicating that something has gone wrong
 or if they were simply unlucky?   For instance
 let's say my department is responsible for responding to customer inquiries within three days
 and I received a complaint that last week they were taking five days to respond. Is this indicative of a problem
 or was it just an atypical week with challenging customers and an increased workload? If I ask them
 they may claim to be unlucky. Therefore
 I need to measure and track the actual situation.   I can create a response time graph
 possibly for each day
 to determine if it is increasing or fluctuating significantly from one day to the next. As we have seen in the S-P-C section
 both scenarios can cause failure. The question is
 where should I set my action level? If someone has only one day of poor performance
 it may not be enough time to make a definitive judgement
 and taking action too soon could be premature. However
 if someone consistently performs poorly for a year
 it would be clear that action needs to be taken. Ideally
 I would be present to observe the situation first and
 rather than relying solely on data. To determine when to take action based solely on performance figures
 I need a clear indicator.   I must observe eight consecutive days of poor or exceptional performance before I can be certain that a change has occurred.   Investigative action is necessary after eight consecutive days. Six or seven days may be attributed to bad luck
 so I must wait and observe. This approach contradicts human nature. After a few consecutive days
 we often feel the need to take action. It is important to note that immediate action is necessary in cases of accidents
 deaths
 or significant financial losses. However
 in most cases
 our natural instinct is to overreact and identify patterns that may not exist. This is a survival mechanism. For instance
 if a bush rustles and a tiger appears
 we assume that every rustle could be a tiger. If someone said they needed to see eight tigers before they would worry
 they would likely be dead. However
 if you flip a coin three times and get heads each time
 you wouldn't assume it's a special coin
 would you? It's important to have sufficient facts and maintain a scientific approach. The answer is eight because that's the point at which it's less than 1% likely to have happened by chance.   If an event is very unlikely and it happens several times in a row
 then you should be suspicious. However
 we're only talking about figures slightly above the average
 which is only 50% unlikely
 and that's where you use this figure of eight. If times
 complaints
 or costs exceed the average eight times in a row
 the probability of this happening by chance is one in 256
 or about half a percent. Therefore
 we can be slightly over 99% certain that something has changed
 and there is less than a 1% chance that it is due to pure bad luck.

We have already established that if an event has a 50/50 chance of happening
 such as something being above average
 you need at least eight occurrences before you can be certain that something has changed. However
 if the event is unlikely
 you need fewer occurrences to detect a change.                     In fact
 the question can be reversed: how unlikely does something have to be before a single occurrence warrants an investigation? What is the warning level at which action and investigation should be taken?   Credit card companies use this calculation to prevent fraud. It may be inconvenient
 but it is necessary. How many complaints would prompt an investigation? How many complaints would prompt an investigation? What is the threshold for going over budget? What is the threshold for accidents or sick days in this area of your business? At what point would you investigate a drop in sales for just one month? It is important to determine trigger points for any measured data.   This information is crucial for effective measurement and analysis. Using the formula of the mean plus or minus 3.14 times the M-M-R (median moving range) can help establish these trigger points.       The following is a column of numbers that represents the number of accidents that occur each month at a railway station. Please note that this does not include train crashes
 but rather minor incidents such as slipping on a patch of spilled coffee. Nowadays
 every little thing is recorded. If I obtain this data from one hundred different railway stations and need to monitor them to determine whether I should investigate a particular station
 I will perform the following calculation: Do not worry
 I will explain. First
 I will calculate the mean
 which is 7.9. Calculating the average is simple: add up all the numbers and divide by the total count. To find the median moving range (M-M-R--)
 arrange the absolute differences between each consecutive number in ascending order and select the  middle number as the MMR. In this case
 the M-M-R is four. Improving this text was straightforward. You can automate the process using a spreadsheet or computer. The median moving range (M-M-R--) is the middle number and represents the typical amount that the number changes from one day to the next. The content of the improved text is as close as possible to the source text
 and no new aspects have been added. A small M-M-R means that even a slight change in the number of accidents will have a significant impact
 while a large M-M-R requires a substantial change in the number of accidents before any concern is warranted. Simple vocabulary has been used to ensure that the text is accessible to a broad
 general audience
 including those with cognitive disabilities
 low reading literacy
 or who are encountering an unknown topic or language. The content of the improved text is as close as possible to the source text
 and no new aspects have been added.   In this example
 the M-M-R is four. Therefore
 the action point is approximately 12.5 above the mean
 as indicated by the formula. To work it out
 find the gaps
 identify the middle-sized one (M-M-R--)
 multiply it by 3.14
 and add it to the mean of 7.9 plus 12 1/2
 which is about 20 1/2. If I hear of 21 accidents

In every process
 there are three factors to consider: quality
 cost
 and time. Quality refers to the standard of the product or service being produced
 while cost refers to the expenses incurred during the process. Time
 on the other hand
 is not simply the number of hours spent on the work
 as this falls under cost.       Rather
 it refers to the leadtime  the amount of time customers have to wait after placing an order. Time can also refer to on-time delivery
 although this is usually affected by variability
 which is more closely related to quality. If tasks vary each time they are performed
 they may need to be repeated
 which can affect on-time delivery. Additionally
 producing things quickly can also be costly
 which I will explain shortly. Therefore
 the root of your processes is quality
 cost
 and time. Unfortunately
 it is impossible to have all three simultaneously.    Reducing quality can lower costs
 but higher quality usually takes longer to produce. Many organizations fail to consider the implications of their desire to provide high-quality products or services at a low cost and with quick turnaround times. Sales representatives or advertising campaigns may promise such outcomes to customers
 but the plan to achieve them often involves putting excessive pressure on employees or assigning a single person to handle multiple tasks. This approach is not sustainable and can lead to negative consequences. Sales representatives or advertising campaigns may promise such outcomes to customers
 but the plan to achieve them often involves putting excessive pressure on employees or assigning a single person to handle multiple tasks. The text appears to be error-free and already meets the desired characteristics. Therefore
 no changes have been made. There is a risk that the mixture may not be optimal. The strongest party will prevail
 or you may end up with a mediocre combination of quality
 cost
 and time. Depending on the market
 it may be more appropriate to aim for maximum quality or maximum speed of delivery
 and make a conscious
 realistic decision to be less good at the other two factors.       In most markets
 there has been a shift from cost to quality and then to time. Until the 1950s
 the main focus was on being the cheapest. However
 quality became an issue and Japanese cars transformed from being cheap to high quality. As a result
 low-quality producers mostly went out of business. Nowadays
 the focus is mainly on time.    Nevertheless
 the Chinese have shifted the emphasis back to cost due to their lower wage costs and economies of scale
 which allows them to produce things incredibly cheaply. Cost is currently a focus due to the ease of shopping around online. However
 this may only be a short-term trend. Research suggests that high-quality producers
 such as German cars or Apple laptops
 gradually gain market share and make more profit per unit
 resulting in the highest overall profit.       The next section will explore the relationship between quality
 cost
 and profit.    We will revisit lead times later. Before that
 consider your organization or market. Is it cost-sensitive? Should you prioritize low costs and prices
 high quality
 or rapid delivery times? What are your competitors doing? Do you need to change your overall strategy?

Does it cost more to make a better product or provide better service? At first sight
 you might think so. However
 there are two parts to the cost of quality. The first is prevention cost
 which increases as quality improves. Prevention cost includes ensuring that things are done well
 such as training
 quality control systems
 good management
 excellent equipment and materials
 and employing skilled personnel. Making everything perfect can be very expensive. In fact
 the cost increases exponentially as you approach perfection. On the other hand
 the cost of poor quality increases as the quality decreases. The more substandard the quality
 the more you have to pay for scrapping
 reworking
 handling customer complaints
 losing contracts
 and even facing legal action. Using cheap materials
 equipment
 and labour may seem like a cost-saving measure
 but it can actually end up costing more in the long run. By combining the two graphs
 we can see the total cost
 which includes both prevention and failure costs. By combining the two graphs
 we can see the total cost
 which includes both prevention and failure costs. By combining the two graphs
 we can see the total cost
 which includes both prevention and failure costs. As shown on the right
 failure costs have decreased to almost zero
 resulting in a lower total cost. Similarly
 on the left
 we were not doing any prevention
 and the only cost was the high failure cost. In the middle
 where the two graphs intersect
 our failure cost equals our prevention cost
 and the total cost is double each of these. Therefore
 the total cost has a U-shape
 with the minimum cost point in the middle. This is the most cost-effective point. I will explain shortly why being here is not desirable. However
 it is worth noting that many organisations fail to recognise this as the most cost-effective point because they do not measure their failure costs. They are acutely aware of their prevention costs
 such as training
 quality control
 and the use of high-quality materials and machinery. There is always pressure to reduce these costs. Budgets for these costs are monitored and cut slightly every year
 which creates pressure to reduce quality to save money. However
 they often overlook the invisible costs of failure. There is no budget for rework or handling complaints
 and the cost of lost customers or wasted staff time is not measured separately. Therefore
 it is essential to aim for the point of minimum cost
 rather than the point of maximum profit. Moving from five to about seven reduces costs and increases profit.     The point of maximum profit is to the right of the middle on a sales volume or price graph. There are four important learning points to take away from this. Choosing low quality is a false economy. The point of minimum costs is not the best place to be
 as it does not yield maximum profit. The point of maximum profit is at about seven. To get there
 it is important to consider how to achieve the desired level of quality without exceeding the cost-benefit ratio. It is not worth trying to reach a quality level of 10. Finally
 the question remains: how can we create this graph in reality? How can we determine our position on the theory?   The answer lies in listing and estimating prevention and failure costs. Investing in better quality
 whether it's people
 equipment
 materials
 or methods
 may prove to be cost-effective. If your failure costs are greater than your

In any company
 there is inevitably some waste. For instance
 during meetings
 how much time is actually productive? Similarly
 how much of your working day is productive? Time is crucial because your biggest expense is likely to be your employees
 which is essentially their time. Additionally
 what proportion of materials are discarded
 and how much does that cost you annually? Expensive resources such as machinery
 laboratories
 and even office space  what is their utilization? Do they use everything all the time
 without wasting space on piles of unused items?   According to a Japanese theory
 there are seven types of waste
 but I prefer to remember them using the word 'downtime'
 which includes an eighth type.   The 'D' in 'downtime' stands for defects
 which can cost you money even if you discover them immediately. The letter O in downtime refers to overproduction
 which occurs when more products are made than necessary. This may happen when items are produced in large batches or for stock
 resulting in the need for storage and inspection
 and tying up capital. The letter W in downtime stands for waiting
 which refers to the time that a product spends waiting
 rather than the time that people spend waiting. This is a half-finished product that may be waiting for a machine to be set up or for other parts in the batch to be worked on. It may also be waiting for a broken machine to be repaired. This waiting time is what distinguishes the short time it takes to make the product from the full lead time that the customer has to wait for it. Remember
 this applies not only to manufacturing but also to services. For example
 waiting for the keys to your rental car is a service
 and the keys are the product you are waiting for. Consider the time spent waiting for them.   The 'N' in downtime refers to unused employee talent
 which is the eighth waste.  Every talented employee is a valuable resource that should not be wasted.   Many organizations fail to tap into the full potential of their employees
 resulting in a significant loss. It is common to find individuals who are disengaged from their work because their employer has failed to motivate them. However
 these same individuals may be highly engaged in their personal lives
 coaching sports teams or engaging in other impressive activities. They have spare energy and potential that work isn't using. Downtime includes transportation
 which is the movement of components
 paper
 or equipment around the site. This can add up if you follow a job from start to final dispatch. Inventory refers to raw materials
 work in progress
 or finished materials ready for shipment. It is important to keep inventory levels low to avoid tying up money and incurring storage costs. It is important to minimize motion to ensure efficient operations. Downtime has two components: inventory and motion. Motion
 on the other hand
 refers to wear and tear on machines and people
 and the risk of accidents and damage. Finally
 the 'E' in downtime represents excessive processing. This includes over-design
 performing operations on parts of the product or service that the customer won't see or care about. Ensure the text is grammatically correct
 uses simple vocabulary
 and is accessible to a broad audience. It is important to avoid adding new content or changing the meaning of the original text. To minimize these issues
 it's important to review every stage of your processes with these factors in mind. Remember to keep sentences

Quality
 cost
 and time are all important factors to consider. However
 lead time is often overlooked.   It is assumed that working harder will make things quicker
 but there is actually a lot of interesting science and theories behind lead time. In this text
 we will explore these concepts further. It is assumed that working harder will make things quicker
 but there is actually a lot of interesting science and theories behind lead time. The easiest way to ensure quick product delivery is to keep them in stock. It may seem like the time it takes to make the product is irrelevant if you have stock available
 but in reality
 stock is often a poor choice. It requires space
 can expire or become damaged
 needs management and checking
 ties up capital
 and can lead to decreased manufacturing efficiency. Keeping stock can limit flexibility and hinder the ability to meet individual customer needs. Accurately forecasting future demand can be challenging
 and not having an item in stock can be problematic. Therefore
 there are hidden costs associated with keeping stock that are often overlooked. Additionally
 services cannot be stored in stock. To ensure timely production
 items ranging from hot pizzas to heart transplants cannot be kept on the shelf. Therefore
 it is important to consider whether items currently kept in stock could be made to order instead. Although this may increase costs
 it could ultimately save money by reducing the need for excess inventory. Whether you are producing for stock or customers
 the question remains: how can you reduce lead times? Why does it take so long to make things?   I used to work in a helicopter factory. It took three years to make a gear wheel for a helicopter gearbox. There were only two weeks of actual metal cutting.The rest of the time was spent transporting stuff around the factory and waiting to start the batch. Often
 there were 10 other batches also waiting for that machine. Then
 waiting for the rest of your big batch. Some batches were 500 parts before finally your part gets made. If you follow the piece of metal
 you will find that it undergoes four main activities. Firstly
 it transfers from one part of the process to the next. Secondly
 it waits in a queue to be operated on. Thirdly
 it waits while the rest of the batch is operated on. Finally
 the only value-added activity in all of this is when it is operated on itself. The same principles apply to an office environment. Therefore
 the lead time could be the time from receiving a planning application to approving or rejecting the new building. It could also refer to the time it takes to get from the airport to the plane
 or to service your car
 or to receive your pizza in a restaurant. Additionally
 it could refer to the time it takes to create a training course like this one. All of these are lead times. Queues now exist as piles of paper on desks
 in email inboxes
 and as waiting lists.   A six-month wait for a doctor is essentially a six-month queue of people. To expedite processes
 there are three methods: firstly
 improve the layout to enable faster movement of items between machines
 people
 or departments; secondly
 use active voice to make instructions clearer; and thirdly
 use verb phrases instead of noun phrases. There may not be significant savings
 but it is worth considering
 particularly for large-scale projects such as car manufacturing.

To streamline any process
 the first step is to consider reducing the number of steps.   This is part of Business Process Reengineering (B-P-R--)
 which involves examining the process and creating a flow diagram. Then
 ask four questions: Can any steps be deleted? Your process may have more steps than you realize. Forms are filled out
 data is entered into computers
 work is assigned to people
 and supervisors check for quality. Consider whether the supervisor's check is necessary
 if the manager can input their own data
 or if the order form needs to be printed.   Another aspect of B-P-R is automating steps through the use of computers
 which can be just as effective as eliminating them. Drawing out the entire process and consulting technical experts can help identify opportunities for automation. Part three of B-P-R involves identifying steps that can be combined and assigning them to a single person. For example
 the person who receives a phone call can also be responsible for resolving the issue. Can a designer also purchase the parts they have designed?   Can a machine operator also pack their work? It is surprising that some lawyers still dictate emails to their secretaries. Consider if there are any opportunities to combine stages in your systems
 which could eliminate a transfer and a queue. Although person two may be faster at completing your job
 it is still likely more efficient for person one to complete all the tasks rather than waiting for person two to finish. Additionally
 the fourth aspect of B-P-R involves identifying steps that can be completed in parallel rather than sequentially
 which can reduce the number of steps and speed up the process. It can save you money to eliminate unnecessary steps or automate them.   Some manufacturing companies have taken this to the extreme with production cells. Volvo conducted an experiment where a team of people built an entire car from start to finish
 instead of using a conventional flow line where the car is passed along to different stations for each step. Although the team was not as fast as specialists in assembling the car
 they saved time on transfer and waiting. The car was worked on continuously
 and the team was dedicated to ensuring quality improvement.   So
 in the idea of a production cell
 a small group of people
 often sitting or standing in a U shape
 and by the way
 they're on the inside the U of tables
 so they can swap around and help each other
 they process a job from the start to the finish. Additionally
 using active voice and simple sentence structures can make the writing clearer and more accessible to a broad audience. It is important to avoid adding new content or changing the meaning of the original text. Setting up teams or cells can increase productivity and efficiency
 especially when assembling similar items. Consider creating cells for family groups of items that need to be assembled.   Although it may not always be efficient for each individual
 sometimes we have to sit through agenda items that do not affect us or require only brief input. However
 this approach is very efficient for the task at hand. Instead of being passed from one desk to another for weeks
 it is dealt with by everyone in a few minutes
 and the job is done. Therefore
 there is a trade-off between individual efficiency and processing jobs within a short lead time. Drawing out and reviewing

In the office dining room
 you receive your soup in 20 seconds. Next
 you move to the meat counter where a nice piece is cut for you in just 3 seconds. After that
 you get your vegetables in 20 seconds. Finally
 one server provides you with both your dessert and drink in 15 and 10 seconds respectively. The payment process takes a minute.   Take a moment to review the diagram. Can you determine how many lunches this setup can serve per hour?This question highlights a crucial and straightforward concept. Avoid getting bogged down with a calculator and numerous calculations. What is the bottleneck of the system? The cashier is the slowest part and can only process 60 people per hour. The number of people you can serve depends on what others are doing. What is causing the delay?   Identifying the bottleneck in your work is crucial. As the manager
 how can you enhance the current setup without investing in additional equipment or personnel? Please note that self-service is not permitted
 and portion control must be maintained.   Remember
 if you concentrate on the bottleneck
 the solution becomes apparent. There is no value in performing tasks faster than the bottleneck;it is a waste of capacity. To improve efficiency
 remove anything that is faster than the bottleneck. This will allow for the redeployment of two people
 resulting in a team of three.   With this reduced team
 they can still serve all the food within the rate of one person per minute
 which is the speed at which they are currently processing orders at the cashier.If you have more demand than one person per minute
 it is better to serve the extra customers by speeding up the bottleneck. Let's suppose you have two cashiers.     Now you can serve a person every 30 seconds
 which is 120 people per hour. By adding just one extra cashier
 you can sell an additional 60 lunches
 which at a cost of $10 per meal
 would result in an extra $600.     This demonstrates the value of working on the bottleneck
 as it frees up the entire system. Consider how this concept could be applied. Finally
 returning to the original setup
 how long will it take to receive your lunch? This question may seem easy
 but it is actually quite challenging. Please give it a try and see if you can come up with the correct answer. You will feel accomplished if you succeed.

Understanding the bottleneck idea and queues is crucial as it applies to any operation
 including yours. There are three perspectives to consider when determining how long it takes to get lunch.   The first is from the viewpoint of the first person in the dining room. They go through all the stages
 and it takes a total of two minutes and 35 seconds to get served. However
 I do not have any concern for them as they are a unique case. What about the rest of the people? The situation is different for them. Therefore
 the second approach to consider is to envision a conveyor belt where everyone moves once per minute.This is the scenario when all the trays slide on rails.     Getting your soup takes 20 seconds
 but you have to wait a whole minute before moving on to the meat because all the trays are waiting for the cashier to serve their person. This means it takes a total of five minutes to get your lunch. Another way to think about this problem is to consider what would happen if there were islands. Imagine visiting the soup island
 enjoying a bowl of soup
 then moving on to the meat island
 and finally reaching the cashier's island.     This analogy highlights the flow of jobs between different offices or departments. It takes only 20 seconds to get a bowl of soup
 allowing three people to be served per minute. They will go to the meat island where only two people are served per minute due to the 30-second time it takes to prepare the meat.   As a result
 the queue at the meat island increases by one person per minute. Two people arrive at the cashier per minute
 but only one person is served per minute
 causing the queue at the cashier to also increase. After 10 minutes
 there are 10 people waiting at the meat island
 resulting in a 5-minute wait to receive their meat. Additionally
 there are 10 people waiting at the cashier
 resulting in another 10-minute wait. In total
 there is a 15-minute wait
 not including serving times. This situation is similar to jobs piling up in the accounts department or people queuing for airport security
 where queues can become very long. Islands are useful for those who want to skip certain food items
 but they can cause uncontrolled queues and delays. Meanwhile
 the conveyor belt system puts pressure on the slowest point to speed up
 which in this case is the cashier. Meanwhile
 the conveyor belt system puts pressure on the slowest point to speed up
 which in this case is the cashier. In a work setting
 management or those being held up could assist. There are three main conclusions from this example. Firstly
 it is easy to determine the system's delivery capacity by identifying the bottleneck
 which in this case is the cashier. And finally
 the text should flow logically and any processes should be described in order. Secondly
 the bottleneck restricts the throughput. Either plan everything to go at that speed or speed up the bottleneck. Investing a little more in the bottleneck can yield significant gains for the entire system.   In our case
 adding a second cashier cost only an extra $10 per hour
 but we served an additional 60 meals
 resulting in a $600 profit. Therefore
 it is essential to take care of the bottleneck. Additionally
 it is crucial to consider the time spent in

In the previous section's lunch line example
 we observed how the cashier was the bottleneck. Identifying and enhancing the cashier's throughput made all the difference. The concept of bottlenecks originated in the 1980 seconds with the publication of Eli Goldratt's book 'The Goal.' Goldratt's book had a significant impact on me
 as well as many others.     His first point is that you cannot produce more than the bottleneck can handle.   Knowing the bottleneck allows you to determine the maximum production capacity
 regardless of other factors in the factory or office. Increasing the bottleneck's capacity will increase the overall system capacity. Therefore
 investing in the bottleneck is worthwhile
 while investing in non-bottleneck areas is a waste. The bottleneck concept leads to many subtle ideas
 and here are three. To avoid work or people piling up in your system
 it is important not to pushing more than the bottleneck can handle. Use this knowledge to schedule the input of work or materials
 as well as what will come out of the system.   If the bottleneck can only process 10 per day
 then the whole system will only produce 10 per day. There cannot be more than 10 items produced
 nor can there be less. If there is a temporary delay downstream
 you can catch up the following day and return to the average of 10.     Another concept that arises from bottlenecks is that time lost at non-bottlenecks is not truly lost because it can be made up as long as it does not affect the bottleneck itself. Ideally
 there should be a buffer of work in front of the bottleneck to ensure that it never runs out of work to do.   Therefore
 upstream issues are irrelevant as the process continues to work on its buffer
 and downstream issues are also irrelevant. Therefore
 management efforts should focus on ensuring that the bottleneck never fails
 always has sufficient work to do
 and is never supplied with defective parts or materials. The third important concept from bottleneck theory is to avoid having operations that require significantly more resources than the bottleneck. This is wasteful because the output will never get through the bottleneck. However
 do not reduce these operations to the same level as the bottleneck
 as this will create two bottlenecks. If the bottleneck is eliminated
 the next lowest production area must be identified .Every system has a bottleneck
 so what is yours?   How do you find your bottleneck? There are three ways to approach this. Firstly
 calculate the required hours for each stage to handle the total throughput and determine if there is enough capacity at each stage. For instance
 if it takes one minute to attach the wheels to the car and 600 cars are produced daily
 do we have 600 minutes
 which is equivalent to 10 hours
 of available personnel at the wheel attachment stage? Secondly
 to identify the bottleneck
 a much simpler method is to locate where the queues are.  If there is a queue for the photocopier
 it may be the bottleneck for your office. Similarly
 if the booking-in clerk has a pile of paper on their desk
 they may be your bottleneck. It is unacceptable if one of your cheapest resources is strangling the output of the whole operation. To identify a bottleneck in a system with multiple products and paths
 examine which items tend to be delayed and locate where they are being held up. This is likely to

Queues are the primary cause of long lead times
 and they build up in front of the bottleneck. When you have enough capacity
 the queue will continue to grow if the number of arrivals exceeds the number of people served.     For example
 if you can serve 10 people in an hour
 and 12 people arrive per hour
 the queue will grow by an average of two people per hour. After 20 hours
 the queue will have 40 people
 which is undesirable. If you can serve 10 people in an hour and only eight people are arriving
 it may seem like there won't be any issues.     However
 the problem arises because the eight people won't arrive evenly spaced. This results in a queue forming when several people arrive at once
 followed by a gap where no one is being served. As a result
 your capacity to serve 10 people per hour is reduced to maybe only seven. The first problem is fluctuating arrival rates
 which also leads to fluctuating service rates. This means that sometimes it takes only four or five minutes to serve a customer
 while at other times it takes seven or eight minutes. The latter group of customers reduces overall output. Additionally
 when a customer only needs five minutes of service
 there will be a gap in which capacity is wasted and employees are idle.   When planning the necessary capacity to provide good customer service
 it is not sufficient to simply calculate the average number per hour. It is important to consider the amount of spare capacity required to handle fluctuations in arrival and service rates. Fortunately
 there is a formula available to determine the necessary extra capacity. The formula states that the queue length is equal to the utilization divided by one minus the utilization. If percentages are used to represent the utilization
 such as saying 75% of the time they are working instead of saying they are three-quarters utilized
 the formula for predicting the queue length will be the utilization divided by 100 minus the utilization. Keep in mind that this is only mathematics
 so there is no need to panic .This formula is easy to use and applies to situations where work arrives at a random rate and you need to serve the customer quickly. For example
 an IT help desk customer service line where it takes seven and a half minutes to deal with the call. On average
 you should be able to serve eight people per hour .If you have 64 calls coming in per hour
 you would expect to need eight people to handle the demand. However
 there will be times when there is no demand
 resulting in wasted capacity and fewer than eight people being served per hour. On average
 64 items arrive per hour
 causing the queue to grow. To prevent this
 spare capacity is necessary. But how much? If we increase the number of workers from eight to ten
 will that be enough? Let's use the formula. First
 we calculate our utilization. With ten people doing the work of eight
 our utilization is only 80%.According to the formula
 the queue length will be U divided by 100 minus U In this case
 it is 80 divided by 100 minus 80
 which equals 20.Therefore
 the queue will have four people. As each person takes about seven minutes to serve
 the waiting time will be approximately 28 minutes. This may be considered too long. Suppose we still receive 64 calls

The basic idea is to avoid producing work that the next person in the chain cannot handle. Buying food from a supermarket or ordering books from the internet are both examples of pull systems because you purchase items at your own pace. The e-mail inbox
 post
 and most factories operate on a fixed delivery rate
 regardless of demand or capacity. The pull system aims to prevent upstream work unless the downstream can handle it. There are two simple ways to accomplish this. The first is to implement an ordering system
 similar to that of a supermarket. The second involves sending a message from one person to another
 requesting either more production or delivery. This message may be in the form of a Kanban card
 which includes the part number and required quantity.   The first person delivers the parts with the card. The second person can then use the card to reorder when they need more parts. Perhaps the card is affixed to the front of the transport box to indicate the parts
 or it is placed at the bottom of the box for easy reordering once all the parts have been used up. The crucial point is that the system cannot contain more items than the number specified on the card or the number of boxes in the system. Person one cannot commence any work until the card has been received. The system is simple and effective
 even when dealing with a large number of contributors. To maintain a manageable workload
 a storage area can be implemented between the two parties. Person two retrieves work from the storage area
 and person one can only produce more work when there is space in the buffer. This ensures that if person two stops for any reason
 person one must also stop. Although this may not sound ideal
 it is an effective way to manage workflow. Remember
 there is no benefit in person one producing more than person two requires. Instead
 person one should assist person two.   This is an interesting and subtle additional advantage of maintaining low stock levels with a pull system.As there is only a small buffer between them
 the two individuals are closely connected
 and they must work more collaboratively. Likewise
 if person one encounters an issue
 person two will run out of materials and will need to check on person one's well-being and offer assistance. Improved teamwork and capacity balancing are achieved through a pull system. For instance
 if someone has more than 20 e-mails in their inbox
 you cannot e-mail them.   In this case
 you can either assist them with their workload or contact a colleague with available inbox space. This approach can be an effective way to work. Why contact someone who is already busy when you have other options? Once the pull system is functioning and buffers are being replenished smoothly
 buffer size can gradually be reduced. To summarize
 why should person one continue producing when person two cannot keep up with the supply rate? Firstly
 it can result in a backlog of work and longer lead times as items have to wait for the next person in line. It is important to have a clear process in place to address these issues.   Secondly
 it can mask underlying issues such as inconsistent quality or high defect rates
 as the next person in line can simply sort through and select the best items. There are two main reasons why having multiple people involved in a

Earlier
 we learned that smaller batches result in less stock or work in progress
 along with other benefits. However
 the reason for batch production is to reduce setup costs.   If it takes an hour to set up the machine and then one part can be produced every second
 how many parts can be produced in a batch?Take a moment to consider before I reveal the answer. It takes an hour to set up the machine
 but once set up
 it can produce one item per second. If you only need one item per month
 then setting up the machine for an hour would produce 12 items
 which is much more efficient   than producing 3600 items.   The production rate is affected by the usage rate
 and storage cost is another factor to consider. For example
 if we're talking about sofas
 it wouldn't be practical to have 3000 of them sitting around. The E-B-Q
 or economic batch quantity
 is a formula that was first developed 100 years ago. It suggests that the optimal batch size is the square root of two times the annual usage
 multiplied by the setup cost
 and divided by the annual storage cost per unit. However
 many people today consider this formula irrelevant. Don't spend too much time learning it. The current belief is not that setting up takes a long time
 requiring a large batch
 but rather questioning why the setup time is so lengthy. When the idea of reducing setup time
 possibly to almost nothing
 is considered
 it alters everything. I recently saw a group of people change a car engine in just 58 seconds on TV. They drove the car into the studio
 uncoupled the old engine
 lifted it out
 replaced it with a new one
 and drove the car out of the studio in under a minute.   This impressive feat raises the question: why does it take an entire hour to change jobs on a milling machine or switch out tools on an injection moder? Filming people during the process reveals that they struggle to locate tools
 often need to retrieve additional equipment
 encounter breakages
 struggle with lifting
 and require a pallet truck. Additionally
 it may take several attempts to achieve the correct adjustment. Streamlining these processes can result in significant time savings during batch changes. You can make smaller batches efficiently
 saving space and lead-time throughout your process. The same principle applies to purchasing. The economic order quantity (E-O-Q--) formula can tell you how many items to buy at once.   As previously mentioned
 the cost of usage
 ordering
 and storage are taken into account. However
 the primary concern should be why a system cannot be implemented to allow for smaller quantities to be ordered and delivered.   Although it may be more expensive to produce
 the savings on shipping and other costs make it a viable option in certain cases. Even in office settings
 work is often batched due to the cost of changeover. Does the person who processes invoices store up a week's worth and then do them all on Fridays? Could they process them daily to ensure suppliers receive payment promptly? Purchasing is often centralized to buy items at a lower cost. However
 this practice can result in hidden costs such as waiting and administration.   Therefore
 reducing set-up time not only saves time but also allows for smaller batches
 reducing lead time and stock holding costs.   Formulas such as E-B-Q and

Quality isn't just the job of the quality department. Ideally
 the person at the top of the company would be obsessed with quality
 seeing it as a way to beat the competition
 get happier customers and reduce costs
 instead of just a pain that means that some things don't get shipped because they fail inspection. And as well as support from the top
 quality should involve everyone in the company. Often
 production people see quality just as inspection and audits to be passed every now and then. Well
 actually
 these production people are the most valuable process improvement asset that you've got if you could motivate them to care about quality. Back in the 1960s
 Professor Ishikawa
 in Japan
 had the idea of a quality circle
 where management would actually ask the workers for help and ideas. This approach was considered radical when it was first introduced
 and it may still be seen as such in some companies. The concept involves assembling a team that includes individuals from all levels and disciplines to work   on a specific problem and find a solution.   This approach recognises that ordinary people
 regardless of their education or position in the hierarchy
 often have the best ideas because they have a deep understanding of what really goes on. They are also motivated by being asked to contribute to a team effort that addresses an important problem. Quality circles were successful in the 1960s
 but have since been forgotten.  Continuous improvement
 or Kaizen
 has replaced them. The idea is that if everyone improves their part of the process by a small percentage
 the entire company will move towards perfection. Toyota implements 5000 improvement suggestions per week. They implement all 5000 suggestions they receive. Their workforce is always brainstorming ways to improve things. While most Western companies have suggestion schemes
 they typically only receive about 20 ideas a year
 half of which are anonymous and rude suggestions about the boss. This is a real problem for continuous improvement in Western companies
 as it can lead to boredom with the process. If you are struggling to maintain regular process improvement meetings
 consider setting up a quality circle to investigate and fix your biggest problem. Provide them with the necessary time
 money
 and empowerment to achieve their goals. By doing so
 you will be surprised at how much they can accomplish and how much they will enjoy it. Therefore
 think about your most significant quality issue and who you could include on the quality circle team.

Two final thoughts. Firstly
 a culture of great quality and service
 which likely results in shorter waiting times
 must come from the top
 even if it requires increased costs. However
 it will be worth it. Sometimes
 you have to spend money before you can save it or sell more and bring in more revenue.This act of faith requires senior managers who understand queue theory and the true cost of poor quality. If needed
 you may be able to persuade them to consider this cost. My second point is tonsure the accuracy of data and personnel before addressing system issues. Reduce work in progress and focus on bottlenecks to implement Lean principles and improve quality.