Control

Control

For the control stage, a 30 day review period was agreed upon, as a means of bedding in the new state, and finishing off some aspects of the kaizen that were unfinished due to resource constraints. This 30 day period will also allow for new problems to emerge, for stakeholders to reflect on the changes, and see if further improvements can be made.

Part of the control phase will be to further educate and empower the operators to take ownership of the process, to improve the process themselves, within the bounds set down by regulatory/ customer requirements.

Documentation and signage were updated, reflecting the new SOP that exists. This new SOP details every step of the process, how each process is done. This has greatly helped to reduce the variation that used exist between operators.

 All relevant paperwork regarding the new layout was also completed.

Cost Saving

Following this improvement project, analysis of the improved process has also been completed, and verified by the Operations Manager. We believe that we have met our initial goals, and as a result, the required cost savings, on an annualised basis, have been achieved.

Conclusion

Overall, this was an enjoyable and interesting project to undertake. It was a great opportunity to learn, to put theory into practice, to work with a great team. It has helped the company to further move along their lean journey, it can be used as an example of what can be achieved by colleagues.

Undertaking change, whether big or small, is never easy. It requires a lot of patience, small steps, lots of encouragement and communication.

Some of the bigger changes proposed or this cell during the analyse plase have been postponed, as they will require a longer timeline to change, due to quality control issues, other business projects taking priority, availability of personnel.

Improve

Improve

The first part of the Kaizen is its organisation. A daily schedule was produced and circulated to the team the week before, so arrangements could be made to reorganise meetings etc.

The first day of the event stated with session on why the improvement was required, how customers requirements have changed, how we need to meet those new requirements. This then moved onto presenting a current state map of the process, discussing what were the problems involved. After lunch, the Lean Six Sigma process was explained, examples shown of how it improves various processes, and how it was applicable to improve the process under review.

Day 2 consisted of brainstorming to get multiple solutions to the problems outline previously.

Once solutions were agreed upon, we moved onto the floor to change the layout, and begin the training to up-skill operators to work the new SOP.

Day 3 continue the changing of the layout and up-skilling.

Day 4 was about putting the new skills to the test in the new layout. Any new problems that emerged due to the changes were solved.

Day 5 was about doing a practice production run, to see could the new production targets be met, and to see if any other changes needed to be made.

New current state

The revised layout, and reduced cycle time, means that operator 1 can now do operations 1-5, while operator 2 does 7-9.

This means that one operator has been freed up to work in other areas of the company.

Figure 11 below shows the improved operator times for the manual processes.

 Fig. 11

Figures 12 & 13 below show the new OEE scores for the process. All stages show an improvement.

Fig 12

Fig 13

The shift pattern also changed, with a 3*8*5 system now in place, with equal numbers on each shift.. This means that output should be equal among shifts.

The new line metrics board, as shown below, is also a success. It has reduced counting time, is easier to update and understand. This has meant that production time per shift has increased. In addition, OEE scores are now displayed, along with a Box Score chart, as shown below.

Fig. 14

Fig. 15

Available working time has increased to 22 hours per day. This was achieved by changing the line metrics board, the shift changeover routine and by staggering break times to make sure bottleneck machine are working as much as possible.

Machine reliability has improved, with inspections now taking place daily, and issues fixed at that time.

Bottleneck throughput has been increased by 10%, not the 15% hoped for.

Yield has improved overall, from 92% to 96%, as compared in figures 9 and 13.

Cycle time has been reduced by approximately 5 seconds.

The average defect rate have been reduced, as can be seen in the OEE scores above. Without further automation, this rate will continue to vary, due to the high manual handling content.

Variation between operators has been reduced, as can be seen in the new process times above.

WIP has approximately halved, due to smaller batches being processed, and the even distribution of work per shift.

Analyse

Here, analysis of the various collected data is done to provide information on where the biggest areas of improvement lie, to brainstorm potential solutions and plan to implement these solutions during the Improve phase.

WIP

WIP needs to be reduced, as too much inventory is tied up at all stages to cover for various issues. Inventory is one of the seven original wastes.

Line Metrics

The line metrics board will be updated, so that the information is easy to read and understand, easy to update, can track output per shift.

Cycle Time

From analysis of the time and motion studies, improving the manual work aspects by standardising the work of various stages can reduce the cycle time, allowing got more pieces to be produced each shift.

Process Yield

Eliminating defects due to the supplier, which are caught in stage 1, will provide the biggest improvement in process yield, while training operators in better manual handling procedures will reduces the in-process defects caught in stage 8.

Layout

Changing the layout will improve the flow of the product through the cell. It will also, in conjunction with takt time calculations, allow for the reassignment of one operator, and the work to be divided more evenly between the remaining two operators. Kanbans will also be used to move product through the line.

OEE

OEE (Overall Equipment Effectiveness) charts were introduced in week 11, as a means to add value to all the primary data collected. Visit www.oee.com for further information.

Figure 7

Figure 7 shows how OEE scores are arrived at, where the losses occur and are assigned.

Plant Operating time is the amount of time the facility is open and available for work.

From here, Planned Shutdown ( schedule maintenance, no products to produce, lunch, breaks) is subtracted to give Planned Production Time.

OEE calculations begin with planned production time, and look at the various losses that occur, with the goal of eliminating these losses. Down time losses are times which stop the machine from working, IE failure, material shortage, changeover time. The remaining time is Operating Time. Operating time as a percentage of planned production time is known as Availability.

This time is reduced by machines not running at maximum capacity, by machine wear and tear, operator inefficiency etc.The leaves us with Net Operating Time. Net Operating Time as a percentage of Operating time is known as Performance.

This is reduced by the time spent making defective products, including parts that require rework.

This leaves us with Full Production Time. Full Production Time as a percentage of Net Operating time is known as our Quality output. This is also our first time yield of the process.

These three numbers (Availability, Performance, Quality) are then multiplied to give the OEE score for the process.

Above information taken from http://www.vorne.com/pdf/fast-guide-to-oee.pdf

                        

Fig. 8

Figure 8 show how the percentages for each stage are rated, and how multiplying them quickly reduces the OEE figure. Quality is the most important of these figures.

Fig. 9

Fig. 10

Figures 9 & 10 show OEE scores for weeks 11 and 12. It shows how the process can change on a weekly basis.

     

Stage 6 is not included in the OEE chart as it is done in another area by a different person.

For week 11, machine reliability/unavailability was a big factor on many of the stages, resulting in the poor scores for availability. Stage 3 was used for some R&D during the week, resulting in no products being processed for a time. This created a domino effect, with parts being unavailable at the next stages for a time also, as no buffer WIP was available to keep machines working.

Performance scores suffered as a result, with operators giving time to fix/ adjust machines, thus targets were missed.

Quality scores are good overall, with stage 1 catching all raw material defects, and stage 8 catching the majority of in process defects.

In addition, these scores highlight the areas with greatest potential for improvement for the Kaizen week.

Following on from figure 5 in the measure stage, and taking OEE scores into account also, stages 3 and 6 are the areas that require greatest improvement, as they are the bottlenecks, with stage 7 running at only 75% during hours worked, and stage 3 running at 79%. In addition, the number of hours worked per day also needs to improve, by reducing setup times, length of handover meetings, staggering break times etc.

Overall, the improvements required are:

• Increase the amount of available working time per day, thus eliminating the need for overtime.
• Improve machine reliability by better maintenance.
• Improving the throughput through the bottleneck processes by 15%.
• Reduce cycle time by 5 seconds.
• Reduce in process defects from 3% to 2%, by better handling of the product.
• Reduce the variation in output between operators by 5% by further training and upskilling.
• Reduce WIP in the line by improving the flow through the line.
• This should all lead to an increase in output by 10%

OEE charts are a very valuable tool to identify areas of improvement in a process. It turns the raw data into easy to understand information, it allows for problem areas to be identified easily, RCA to begin on these areas, solutions to be implemented, and the results identifiable on updated OEE scores.

Measure

Measure

The measurement stage kicked off at the same time as the define stage, as data had to be collected over a number of weeks. Getting one week of data is not enough for a baseline, as it may be a good week or a bad week. Getting the average performance over a few weeks gives a more reliable indicator of performance.

Getting the data down on paper is a very valuable exercise, as it allows the analysis stage to be carried out.
It allows us to gather information on what is good and bad about the process, the machinery availability, first time yield and defect rate, where the bottlenecks are, where improvements can be made, where the process needs to be changed, etc.

Current state

Fig 4

Figure 4 shows the current state of the process, how long each manual and automatic part takes. It is the average manual times that are shown here. The data was  gathered using time and motion studies. The important aspect here is that operators keep to their normal routine while being timed, otherwise the study is distorted.

At all stages, there is wide variation between operators, so reducing this variation will be key, and will be looked at during the Analyse phase. This improvement will, hopefully, lead to cycle time reduction, and thus greater output per hour, reaching weekly production targets.

Stages 7 and 3 are the bottleneck in the process, with 6 also being an issue, as it can stop production on 7 if delayed. This is due to the fact it is not processed in the cell, so its output depends on other people. Getting a more predictable output here is key, as it allow downstream process to work better, and smaller batches to be processed.

Stages 1 and 2 are automatic processes, with regular inspection required at stage 1 to monitor and compare output against specification.

Stage 3 controls the throughput of stages 4 and 5, as it is slower than them. Hence the requirement to keep it working as much as possible.

Stage 6 takes place away from the line.

Stage 7 is the key to output, and needs to be the most productive stage. However, this is not the case, and will be a major focus of the Kaizen.

Stages 8 and 9 are quick and simple steps. However, they are processed in large batches. This needs to change to improve the flow of product through the line.

 

 

 

 

Shift patterns

Fig 5

Figure 5 shows the number of people normally available during various period of each day. This uneven distribution is another reason the line is unbalanced, as work is allowed to build up during the day at various stages so as to keep the evening and night shift busy. This is another area to be adressed later on in the project.

Defects

Fig 6

Figure 6 is a pareto analysis of the defects, and the stage at which they occur.  Stage 1 catches the raw material defects, while stage 8 catches most of the in-process defects. These will be the focus during the analyse and improve stage.

WIP

Currently there are, on average, approximately 12000 pieces of WIP in the line. This ties in with figure 5 above, as quantities are built up ahead of the later shifts; and the current state above, as WIP covers over a few problems.

Line Metrics Board

The layout of the line metrics board needs adjusting. Currently, updates are posted about the progress of each job through the stage, with WIP and finished goods recorded. This is slow to update every day, hard for other people to analyse, limited in giving relevant line information. As a result, a new board layout will be trialled during the analyse stage.

Many issues came to light during the data gathering stage, from environmental issues like noise and poor lighting, H & S issues like adjusting machines during calibration, high defect levels leading to quality issues, shift pattern issues, supply chain management, etc

Gathering all the stakeholders into a room and communicating these issues had a very beneficial outcome. People got a holistic view of the process, realised how problems can affect many areas, and proposed various solutions.

Resistance to change is a feature of this improvement  process. This is very natural, as people are creatures of habit, thus change is viewed with suspicion. Understanding the causes of the resistance is vital to ensure the success of this project. 

To overcome this, communication at all levels is vital, visible management and leadership is required, feedback is continuously sought, a workable schedule will be implemented, and upskilling and training will also be required.

Introduction and Define

Introduction

My name is Hugh O Donoghue, and I am a part time student at IT Sligo, currently doing the Certificate in Lean Sigma Quality(Green Belt).

My full time position is as Production Engineer with a medical devices company. Currently I am working on improving various production lines, one of which is the basis for this project.

The main focus of this project is to balance the line, so as to meet customer requirements during the normal working hours, improve the flow of the product through the process, reduce in-process WIP, reduce batch sizes and standardise the work.

During the course of my investigation, I will be getting an understanding of what happens at each stage of the process, the breakdown of the work content of each step of the process, and from here analysis the data and doing calculations such as cycle time and takt time, identifying areas of improvement and implementing solutions.

The process contains 9 steps, each of which contains areas of potential improvement. These may include, but may not be limited to: SMED, operator variation reduction through standardised work, kanbans, Pareto, Poke Yoke, SIPOC.

Define

The initial meeting with the Operations Manager began with the knowledge that the product does not flow through the process.  Further to this, production on this line has increased due to customer demand. As a result, production is not meeting customer demand within normal working hours, resulting, resulting in overtime being required.

From here, we agreed to collect data over a number of weeks so as to investigate why the product does not flow smoothly, and to have some evidence available when a meeting that included more stakeholders in the process was organised.

Currently, of the nine stages in the process, two stages are processed in large batches, as they are long processes.  As a result, some of the other processes downstream of them are also process in large batch sizes. At the conclusion of this project, the goal is that the line will be balanced, so that the product flows in small batch sizes where possible.

Stakeholders

The Operations Manager, Production Engineer, Value Stream Leader, Quality Engineer, Line Engineer, Line Technician, Health and Safety Coordinator, Financial Controller and production operators all have an interest in the reorganisation of the line. Production operators from the various shifts are involved to give a hands on perspective of the work, and to highlight any issues that are applicable to their work time.

Fig 1

In the beginning, there was no scheduled meetings with stakeholders, just informal conversations to gather information and keep the stakeholders up to date.
As the project progresses, a more formal meeting schedule will be put in place.

Value Stream Map

Gathering of initial data produced the process map below. It highlights areas that require further investigation, the bottleneck processes that are stage 7 & 3, idenifies whether processes are manual or automatic. Operator 1 is responsible for steps 1-4, operator 2 is responsible for 5 & 7, while operator 3 is repsonsible for operations 8 & 9.

Fig 2

Project charter

Developing a project charter for a project is a great idea, as it gives a focus to the project, its outlines why the project is required, what the goals of the project are, it gives all relevant data on one page, it gives a timeline. It gives ownership and responsibility to team members.

For me, developing this charter was a great way to get an introduction to this process, to get to know the stakeholders and understand why it was required by the company.

VOC

VOC analysis was carried out when developing the project charter. Lots of information was gathered as to why the process was underperforming, and this provided a roadmap of the areas to focus on later in the project. Much of this deeper information was unknown to some of the stakeholders, due to lack of interaction with the operators.
It was an important part of the project, to get all the relevant information to the stakeholders. It also highlighted tht communication was an area that needed improvement.

Project schedule

In the beginning, there was no fixed schedule, the main goal was just to start collecting data about the performance of the process. From here, the schedule will be influenced by customer orders, overall business performance, stakeholder availability, etc

The DMAIC process will be run in parallel at different times, as measuring will take place over the duration of the project, initially to get baseline data, then tracking the output after changes are made, then ongoing monitoring to make sure the new standards are still being followed.

Fig 3

Regression back to the old state after a while is one of the big problems associated with any change project.

Key Factors

Managing the change process. Resistance to change is very normal. People are used to their work routine. However, these routines need to change as part of the continuous improvement process.

Communication. All stakeholders need to be kept informed of what is happening during the project, why changes need to be made, what data has been gathered,what changes will be made. Communication will be by email, noticeboard, meetings, informal conversation.

Upskilling and training. Getting to a new improved status requires training, as a new standard work has to become the new normal.