Amit Simon QX

Exponentially weighted moving avg (EWMA) and control charts are both statistical tools used for monitoring process performance. However, they differ in their approach and applications. EWMA is a time-series analysis which assigns a weight to each data point and gives greater weight to more recent data. Recent data points have greater influence on avg than older ones. On the other hand, control charts plot the process data over time and indicate the upper and lower limits based on process variation. outliers require further investigation and corrective action. Example: Suppose a manufacturing company produces car batteries and uses an EWMA chart to monitor the average voltage of the batteries over time. The company has set a target voltage of 12 volts, and the control limits on the chart are +/- 0.5 volts. Advantages: The EWMA chart can detect small changes in the average voltage, such as a gradual increase or decrease, which could indicate a problem with the production process. The chart can be easily customized to adjust the sensitivity to changes in voltage. For example, a smaller lambda value could be used to make the chart more sensitive to small changes, while a larger lambda value could be used to reduce the impact of outliers and noise. The chart can be easily interpreted by the operators on the production floor, who can quickly identify when the average voltage is trending towards the control limits. Disadvantages: The choice of lambda can impact the results of the chart, and different values may be more appropriate for different situations. This requires some level of subjectivity and expertise. The chart may not react quickly to sudden changes in the process, such as a large increase or decrease in voltage caused by a specific issue.

In this scenario, I think, the most relevant chart to check process stability is the Xbar-R chart, because an Xbar-R chart is used to plot the process average (Xbar) and variability (R) when subgroup size is constant (i.e. 8 bottles per shift). This is also due to the small sample size and continuous form of data. Both these factors are also ideal for an Xbar-R chart. an I-MR chart is used for only one observation per subgroup. Since QC team checks 8 bottles every shift, I-MR chart is not the best choice. An Xbar-S chart uses (S) standard deviation instead of (R) range. However, Xbar-S should not be used as the data is not normally distributed. Key use of X-Bar R chart 1. Data must be continous 2. Sample size (subgroup) must be at least two and should not be more than 10 to 12 3. Sample size cannot vary 4. Data should be collected in a random manner 5. at least 20 to 25 sample groups must be collected to ensure adequate measure of process variation has been taken. Here is a chart plotted with the given example. Measurements are random.

A Job Breakdown Sheet (JBS) is a document that outlines the steps required to complete a specific job or process. It is a tool commonly used in manufacturing and production environments to improve efficiency and productivity. The potential advantages of using a JBS include: Improved clarity and consistency: By breaking down a job into smaller steps, a JBS provides clear instructions and ensures that all workers are following the same process. Reduced errors and defects: By specifying each step in a process, a JBS can help identify potential sources of errors and defects, allowing them to be addressed before they cause problems. Increased efficiency: By breaking down a job into smaller components, a JBS can help streamline the process, reducing wasted time and effort. However, there are also potential disadvantages to using a JBS, including: Increased complexity: If a JBS is overly complex, it can be difficult to follow and may actually decrease productivity. Lack of flexibility: If a JBS is too rigid, it may not allow for deviation from the process, preventing workers from adapting to changing circumstances or making improvements. When integrated into the broader Lean Six Sigma DMAIC framework, a JBS can be used as part of the Define phase to clearly outline the process being improved. It can also be used in the Measure phase to identify potential sources of variation, and in the Analyze phase to identify and address root causes of problems. Finally, a JBS can be used in the Improve phase to test and implement process improvements, and in the Control phase to ensure that improvements are sustained over time.

The concept of six sigma is all about reducing variation and enabling a consistent output/outcome. This means all factors that could bring in variation are tightly controlled. For this to happen, of course there was a prior plan on how to keep things under control, which is known as a control plan. This includes a comprehensive risk assessment, a detailed thought through assessment on what all could go wrong or what all could fail. While the system delivers a controlled output due to such control plan, there always are possibilities that process may run out of control, producing defective output/outcome. The purpose of OCAP is to identify the root cause(s), take corrective actions to eliminate the causes(s), and prevent the problem from recurring in the future. This can be achieved using quality tools and techniques such as statistical process control, Ishikawa diagrams, process flowcharts, Pareto charts to analyse the data and identify the root causes of the problem. A diligent and consistent adherence of this method leads to a situation where the possibility of failures keep reducing and results in continual improvement. To take the best advantage of OCAP, cross functional teams must be involved and the results must be shared as best practices across the organisation. Learning from other function's OCAPs must be considered too. One important point is to consider the review of the OCAP while changes are planned in the system or processes. While you build a tightly controlled system using years of knowledge, a badly manage change can make the process go out of control.

EVOP is a process or technique of systematic experimentation. An experimental design is established and conducted through a series of phases and cycles, through systematic changes in operating conditions. When we find a factor to be significant, we reset the operating conditions for that factor. Then, we conduct the experiment again. The process continues until no further gain is achieved. EVOP is suitable when a) There is a variation in the process performance over time b) there are 2 to 3 major contributing factors or process variables c) System has more number of product performance conditions Disadvantages of EVOP a) May not always find the absolute optimum, but rather a local optimum b) takes more time to reach optimal settings compared to DOE c) more likelihood of errors as it requires more experimental runs than DOE Example EVOP can be used to optimise a process in a manufacturing plant, where the output required is the number of defect free products. This can be achieved by variying the process parameters like temperature, pressure and time of a critical step in the process. The initial trial is conducted to establish the starting conditions for the process. The EVOP cycle involves making small changes to the process parameters such as increasing the temp by 1 degree and monitoring the impact on the output. The results are analysed and the process is repeated until the optimal conditions are achived. The optimal settings are then implemented in the production process EVOP process steps Once the process is identified 1. Determine the parameters that will be varied during the experiment, based on their potential to impact the output 2. The initial trial is conducted to establish the starting conditions. This should be based on process history and prior knowledge of the process 3. Implement the EVOP cycle making small, incremental changes to the parameters and monitoring the impact on process output. 4. Analyse results to find the optimum settings 5. Once found, implement the optimum settings in the production process

Workload balancing refers to distributing the worload equally among employees or resources to ensure there is no overload and productivity is optimum. It is a method to analyse the workload of each resource and distribute as a balanced workload. In manufacturing industry, this is done through production planning and scheduling, considering the capacity of production resources, including availability of machines and skills of the workforce. In service industry, workload balancing is done using staffing scheduling, where the manager must balance the number of staff needed to meet the current demand. Aspects to consider in manufacturing industry are production time, capacity, machine downtime, maintenance schedules etc. Those to consider in service industry are customer demand, staffing, staff availability etc. AI and machine learning can help with workload balancing both in manufacturing and service industries. AI algorithms can analyze historical data and provide recommendations to optimize production planning and scheduling. In service sector, AI can help with staffing planning and scheduling. Machine learning can be used to analyse employee performance data and workload data to identify patterns, and recommend work load balancing strategies. By using technologies such as AI and machine learning, companies can optimize their work load balancing strategies and achieve better results.

I think in this scenario DPMO would serve as a better capacity metric. DPU is defects per unit however DPMO gives a value of defects per million opportunities. Comparing defects and defectives, where the latter is where the entire unit is not acceptable, calculations on defects are more precise. Dpmo provides a mesure on capability of the process to generate defects which can be used to calculate sigma level. So in case of a discrete data dpmo will be a better process capability metric. Eg soft drink bottles considering the cap, label and bottle diameter, the capability of the process can be better measured using dpmo

5 why analysis helps to understand the underlying root cause due to which a problem is occurring, Is/Is not analysis can be used for root cause identification however it can also be used to define a complex problem. The Is/ Is not matrix helps to separate the factors that are not the problem. This helps to narrow down the scope of the problem and focus on the problem area. In a why why analysis each cause is further studied for factors that make the cause occur. Is/Is not analysis on other hand is based on getting to the most relevant factor by theory of elimination. Of course the elimination has to be done based on certain validation. an example is attached.

The 8 wastes described in the Lean philosophy very aptly define factors that affect the capability of a system to reach maximum efficiency. Overproduction is where an output is produced more or sooner than required. Over processing is unnecessary additional work is done which adds no value to the output. Manufacturing: Overproduction 1) The assembly line of a mobile phone manufacturing company has an imbalance. Phase n is unable to handle the volume produced by phase n-1. This results in inventory waste at phase n due to an overproduction at phase n-1 2) A restaurant cooking amount of food which is more than volume of customers expected. The remaining food goes to waste Over processing 1) A hand sanitizer company adds a fancy package in which the bottle is sold. Customers do not care about how fancy the package is for a product like hand sanitizer 2) A desk calendar manufacturer uses an extremely expensive quality of chart paper which increases the cost of the product SERVICE INDUSTRY Overproduction 1) Recruiting more than necessary people without a scope to place them in operations 2) Generating more sales leads that the operations has the capacity to handle Overprocessing 1) Creating a customer dashboard (MIS) with additional unnecessary data 2) A call center asking too many questions, questions that are not going to be used for the solution. HOW TO HANDLE THESE WASTES Overproduction can be handled by balancing the line by techniques like Just in time. the advantage of pull system can be taken Overprocessing can be managed by understanding the customer perception of value and calibrate the production accordingly. Value stream mapping can help