Case Study Creative Lighting Company
Introduction This case study has been developed to further explain the application of Global 8D Problem Solving (G8D), in particular it focuses on steps D3 to D7. As a reminder the steps in the G8D process are defined as: - D0 Prepare for the Global 8D Process D1 Establish the Team D2 Describe the Problem D3 Develop the Interim Containment Action (ICA) D4 Define and Verify Root Cause and Escape Point D5 Implement and Validate Permanent Corrective Actions (PCAs) for Root Cause and Escape Point D6 Implement and Validate Permanent Corrective Actions (PCAs) D7 Prevent Recurrence D8 Recognise Team and Individual Contributions This case study will also cover specifically the use of the following tools and methodologies: - Is/Is Not worksheet Differences and Changes worksheet Possible Causes worksheet Possible Root Causes worksheet Givens and Wants process Decision Making worksheet Verification testing Validation testing Background This case study has been developed around real events at a lighting design and manufacturing company that we will call The Creative Lighting Company. The company design and manufacture a range of domestic, utility, industrial and retail lighting products. They are a relatively small company based in the UK with customers all over Europe. This case study is used to provide a working example of how the Global 8D problem solving process has been applied to what appears to be a simple problem in their paint plant. Scenario The Creative Lighting Company are making metal lamp shades for the domestic market. These are fabricated from two spinnings which are press fitted together and then painted using a robotic powder coat paint plant in several colours (spinning is a process for making cylindrical forms over a pre-shaped former). These are then assembled with a lamp holder, metal reflector and mounting arrangement for ceiling and/or wall mounted application. 2
Some of the blue finished product has a defect, paint runs on the top flat. This has been occurring over the past few weeks. Top spinning (mounting flange) Flat Lower spinning (reflector and lamp housing) The paint plant is a powder coating plant. Powder is positively charged through robotically operated spray guns that are mounted in shuttle booths, one for each colour and one for special colours. The product is manually hung on the conveyor (that runs continuously) using specially designed hooks. After each paint run the hooks are dipped in a solution to remove the old paint. The conveyor is negatively charged to attract the paint and the metal hook serves to provide that electrical connection. The following diagram gives a schematic of the paint plant area for information. Loading and unloading area Paint spray booth shuttle system for colour change Track for hanging product for painting, runs at a set speed continuously Drying Paint Plant Schematic Diagram Then Creative Lighting Company had been making this lamp shade design for over 2 years when they ran into this problem of a paint run on some of the finished product in one particular colour, blue. As they do not make for stock, only to order for their retail outlet customers, this has an immediate impact on their ability to meet current demand. 3
After several attempts to resolve the problem themselves the paint plant team call in help from the Process Engineer. She decides that this requires a more structured approach and employs the G8D process to ensure they get to root cause and prevent this type of problem occurring again. The Problem Solving Effort At this stage the paint plant team have instigated a 100% inspection to remove any faulty blue lamp shades. This is delaying the change over to the next colour as they are having to paint more blues than needed to make up for the number being rejected at end of line inspection. The Process Engineer has been asked the lead the G8D Team in their problem solving efforts and to ensure that a customer representative is also made available to the team. The team has been formed and includes members from Quality, paint plant, the Process Engineer, Maintenance and from assembly (the internal customer). We join the problem solving effort at D2 Define the Problem. D2 Define the Problem The team leader confirmed the purpose of the meeting which was to establish what had been done so far and to get to a clear problem statement. The team discussed what had been done so far which included: - Use a more experienced operator to load the paint line Increased the drying time by 30 seconds (Speed of conveyor) 100% inspection of primer coated product before finish coat 100% inspection of finished product before moving to packaging ready for transporting to the assembly area Based on the facts available to them, the team then developed the problem description starting with the problem statement what is wrong with what. Problem Statement: Some of the blue finished product has a defect, paint runs on the top flat. This is what the G8D Team have been able to establish as a result of their initial investigations. Having established a clear problem statement, the team now need to develop a problem description. To do this the team used the Is/Is Not worksheet and completed it as follows. 4
Is/Is Not worksheet IS IS NOT (But could be) Data / information needed What Object Domestic lamp shade finished in blue On any other colour of domestic lamp What defect Paint run Blister, low paint cover, inclusions Where on object The top flat Anywhere else Where first seen After top coat applied Before top coat Where observed first After the top coat in the paint plant Any other time Where seen since Only in the paint plant Anywhere else When observed first Three weeks ago Over three weeks ago What since pattern No data Get data from the plant How affected many 20% of the blue N0 more or no less than What size No data Get data from the plant Defects object per No data Get data from the plant Trend Stable Rising or falling percentage The team then went and gathered further information from the paint plant and reported that:- What pattern since? = Random What size? = 2mm by 15 mm consistently Defects per object? = One per finished lampshade The team are now able to develop a problem description for the lamp shade as based upon the information they were able to establish about the problem Problem Description: 20% of the blue finished lampshades are consistently rejected for paint runs on the top flat. Defects are not seen before the finish coat is applied. 5
D3 Develop the Interim Containment Action (ICA) Having established the Emergency Response Action (ICA) as 100% inspection the team looked into various options, considering the following: - Continue 100% inspection loosing 20% of finished blue product Stop producing the blue product Get the blue finished by an existing subcontract paint supplier Rework and recycle all rejects through paint plant Paint the blue product using a wet backed booth normally used for prototype and development work The following was recorded by the team as part of their decision making process: Possible ERA For Against Rank Continue 100% inspection loosing 20% No other action Loss of 20% 2 of finished blue product required Stop producing the blue product Will stop the problem Will lose customers quickly 5 Get the blue finished by an existing subcontract paint supplier Rework and recycle all rejects through paint plant Paint the blue product using a wet backed booth normally used for prototype and development work Will meet customer need. Will buy time to solve the problem. Additional product can be achieved. Keeps the problem in-house Can do all needed and any rework of old rejects Additional cost of 1.50 per item gives a loss on all blue sales Needs to be done on overtime as plant is at capacity right now Needs skilled labour from development dep t. Paint not as heat resistant as powder coating 4 1 3 Having considered the options the team decided to recommend to rework the blue rejects in overtime as the labour could be used on all other colours and only needs to be in attendance for the first 20 minutes of the process for loading onto the paint plant as they can be removed by the day shift on the following day. D4 Define and Verify Root Cause and Escape Point The G8D team had 2 meetings to determine possible causes and then to establish the possible root causes. Possible Causes: Referring to the Is/Is Not worksheet, the team first used the Differences and Changes worksheet to record what was different in, on or around the IS when compared to the IS NOT E.g. What is different on or around the (is) blue when compared with (is not) the black, white or grey finished lamp shades? 6
The team established a number of differences and these were recorded on the Differences and Changes worksheet. Different paint colour Different paint thickness specification (heavier coat) Different paint supplier Different paint powder coat booth for each colour Longer hooks used for blue finish Then we need to ask the question what has changed in, on or around the IS. Something must have changed or the problem would have always been there. This must be factual so it s OK to generate a question log or ask a team member to go and find out. A useful tool that can be used to record this information is a time line. A time line is used to record all events in and around the problem. The team developed their time line having gathered all the relevant details. Time Line for Paint Plant New hooks New paint batch Plant cleaned New electrical earth installed New staff New dryer Roof leak over plant 15/2 21/2 2/3 11/3 18/3 1/4 12/4 This time line clearly shows events that the team considered relevant to the problem with the blue lamp shades starting with a new supply of hooks on the 15 th February to the new electrical earth installation on the 12 th April. This can be added to later in the process as more information and further understanding is achieved. Having this and previous information the G8D team can now complete the differences and changes worksheet The relevant section of this completed worksheet is shown below. 7
G8D Worksheet - Difference and Changes For each entry in the IS/IS NOT Worksheet, determine the difference between the "IS" and the "IS NOT". Then determine what has changed in, on or around the IS and record the date of that change. Differences Changes Dates Paint colour, Paint thickness (thicker on the blue), paint supplier different for blue only, specific booth for each colour, longer hooks used for blue paint due to paint thickness New batch of hooks for all colours, New staff on paint plant, New batch of blue paint, New batch of hooks supplied 15/02/2012 21/02/2012 2/03/2012 15/02/12 Lamp shade hangs lower due to longer hooks used to ensure paint thickness application The longer hooks were introduced with the new paint specification for blue in late December last year. 22/12/2011 This shows the differences and changes around the problem and is driven from the information gathered and the entries on the Is/Is Not worksheet. The team then brainstormed possible causes of (problem description) 20% of the blue finished lampshades are consistently rejected for paint runs on the top flat. Defects are not seen before the finish coat is applied. Possible causes were then brainstormed by the team and they recorded the following:- Untrained labour Poor paint quality Bad connection to earth Clean-down during maintenance caused excess dirt Water got into plant after roof leak Having determined possible causes the team then developed a causal theory for each possible cause and entered onto the Possible Cause worksheet seen on the next page. NOTE: again this worksheet aligns with the previous 2 as the process is driven by the differences and changes around the IS when compared to the IS NOT. Therefore the causal theories are developed in line also. 8
G8D worksheet - Possible Causes Develop rounded causal theories for each possible cause. First brainstorm the causes and develop a rounded statement for each possible cause A) Untrained labour have hung the lamp shades incorrectly leading to incorrect angle and paint runs during the curing process B) Untrained labour touched some of the lamps during hanging. The grease deposited prevented proper curing of the paint resulting in runs C) New batch of paint was of poor quality leading to unsatisfactory curing leading to paint runs D) New hooks not fitting properly causing a poor earth and leading to paint runs due to poor adhesion and curing properties E) Hooks for blue paint finish are longer than normal and are not fully immersed during cleaning. This leads to partial removal of old paint leading to poor earth connection F) Clean-down during maintenance caused excess dirt in the plant leading to paint defects NOTE: This only shows the completed section of the worksheet NEW TEAM MEMBER: having developed the possible cause theories the team decided that they needed some additional expertise in the team and requested that an engineers from production engineering and maintenance attended the meeting to help with the next stage of the process. Establish Possible Root Causes Using the Possible Root Causes Worksheet the extended team completed the worksheet as detailed on the following page. 9
G8D Worksheet - Possible Root Causes For each causal theory determine if the cause fully explains why the problem exists with the "IS" and not the "IS NOT". Use a PLUS sign (+) to confirm, a "?" if not known and stop once a no or NEGATIVE sign (-) is entered A B C D E F G H I J K N N N N N NOTE: the possible root causes worksheet has been shrunk for the purposes of this case study. Normally this would align to the Is/Is Not and other worksheets. 10
The outcome of this exercise showed that possible cause E - Hooks for blue paint finish are longer than normal and are not fully immersed during cleaning. This leads to partial removal of old paint leading to poor earth connection to be the most likely cause. Verify Possible Root Causes The team reconvened and determined that the verification activity would be carried out in four stages: - 1. Inspect all the longer hooks in the paint plant and check for paint residue after the cleaning process 2. Examine the cleaning process to determine if the theory could be proven regarding cleaning 3. Select suspect hooks and run a trial through the paint plant to see if all the blue lamp shades developed the paint run after curing 4. Use the same hooks fully cleaned to see if the problem was not repeated This was carried out later that day and the theory was proven and they were able to switch the problem on and off again. The G8D Team now had a verified root cause for the problem. Define and Verify Escape Point The team discussed with the paint plant the hook cleaning process and agreed that the depth of the cleaning tank was OK provided that the cleaning fluid was kept topped up. Anything less that 90% full meant that the longer hooks would not be fully immersed in the cleaning fluid. The team ran a further experiment to determine how long before the cleaning tank fluid level dropped below 90% but within the minimum level specified in the Standard Operating Procedure. They concluded that this would take 6 to 8 working days to get below the 90% and a further 12 days to get to the minimum level based on current activity levels. This was determined as the verified escape point. D5 - Implement and Validate Permanent Corrective Actions (PCAs) for Root Cause and Escape Point Having established the verified root cause of the problem and escape point, the team met to discuss what would be the most effective permanent corrective action and discussed and agreed the following options: - 1. Design an alternative hook that would fall within the minimum cleaning fluid level 2. Increase the size of the cleaning tank to facilitate the longer hooks 3. Increase the frequency of cleaning tank cleaning and top up 4. Use an alternative cleaning method for the longer hooks (shot blasting or sub-contract cleaning) 5. Use disposable hooks 6. Develop the paint plant (spray booth) to enable the standard hooks to be used on the blue finish 7. Change the SOP for the cleaning tank to revise the minimum level to 90% 11
Each of these was investigated by the team and other members of the company and the following was reported back to the team by the team leader and the production engineering department. Proposed PCA Complexity Time to Cost to Product Disruption to Comments/ implement (Days) implement ( ) on-cost ( ) production concerns 1. Alternative hook H 20 25 100 Nil None Concerns over design paint finish thickness 2. Larger tank L 10 1,500 Nil 1 day 3.Inceased frequency of L 2 Nil 0.03 Nil Concerns over top up and cleaning of tank cost of cleaning fluid at 60 per refill 4. Alternative cleaning method shot blasting M 14 Nil 0.02 Nil Concerns over life of hooks reduced to 3 weeks from 6 months L 10 Nil 0.07 Nil Reliability and Alternative cleaning method sub-contract turn round times 5. Disposable hooks L 14 300 0.36 Nil Cost of hooks and need to stock 6. Paint plant H 30-40 2,500 3,500 Nil 1 3 days Not guaranteed development outcome 7. Change SOP for L 1 Nil Nil Nil cleaning tank operation KE Complexity: L = Low; M = Medium; H = High The team leader had reported to management the success so far with potentially solving the problem and was provided with a number of criteria that the final solution needed to meet: - GIVENS (Must meet these):- Must solve the problem completely Must not disrupt production more than 1 day Maximum cost to implement of 2,000 Maximum on-cost to product to be 0.02 Time to implement must be within 10 working days WANTS (Preferred if possible): - 500 to implement (5)* Zero product on-cost (10)* Implement within 5 days (5)* *The number in brackets is a score of how important these are to the business. The team used the decision making worksheets to establish which ERA proposal to take forward. This activity is summarised in the following table. 12
Decision Making Worksheet Summary Proposed PCA Failed/met Givens Criteria 1. Alternative hook design Failed - time to implement 2. Larger tank Met all givens 3.Inceased frequency of top up and cleaning of tank Failed product on-cost 4. Alternative cleaning method - shot blasting Failed time to implement 4. Alternative cleaning method sub-contract Failed product on-cost 5. Disposable hooks Failed time to implement and product oncost 6. Paint plant development Failed time and cost to implement and product on-cost 7. Change SOP for cleaning tank operation Met all givens Given the failure of all but 2 proposed ERA s, the team agreed to proceed with: - Installing a larger cleaning tank to ensure all hook sizes would be fully immersed and therefore properly cleaned (ERA for root cause) Change the SOP for the cleaning tank to include a regular depth check twice for every shift (ERA for escape point) D6 Implement and Validate PCA s The team drew up a full plan and included all the necessary disciplines from the company in the meeting. These included production engineering, health and safety, maintenance and process engineering in the form of the team leader. The validation was carried out in the following ways: - For root cause - Monitor paint quality by having a daily report on all quality issues to ensure problem remained solved Earth testing kit provided for sampling all hook connectivity every hour which was recorded and reported each day For escape point Developed an operator level check graph that was monitored by the paint shop supervisor and measured Cleaning fluid level twice a day Cleaning fluid cleanliness at the beginning of each shift Sample visual check of each type of cleaned hook The SOP was up-dated to include this and the operators were trained 13
D7 Prevent Recurrence The extended team came together to determine the root cause of the root cause for both the problem and the escape point. It was discovered that the cleaning tank was not considered as part of the induction process and that the operators were not aware of the importance of using only completely clean hooks. The team recommended that: - 1. All new product introduction should be signed off by process engineering as well as production engineering and production management 2. All operators were trained on the principles of powder coating and the key factors to be monitored 3. The induction process for new operators in the paint plant to include a training period on paint plant operation and/or cleaning tank operation NOTE: For further information on the use of all the worksheets and methodologies, please see the Global 8D Problem Solving with Results workbook. Graham Cripps ResultsResults.co Ltd 14