iSixSigma

AVY

Forum Replies Created

Forum Replies Created

Viewing 64 posts - 1 through 64 (of 64 total)
  • Author
    Posts
  • #132415

    AVY
    Participant

    Ashwin,
    It is advisable to do MSA- Gauge R&R for all gauges.
    But if you have a time and constraint you can prioritise based on the criticality of the parameter that is measured by the gauge. I believe all the parameters must have been identified in the Control Plan. You may first do MSA of gauges that are used to measure parameters which are categorised as Critical characteristics (CC) and then Significant Characteristics (SC) in your control plan. ( The frequency of doing Gauge  R&R for the respective gauge, also forms a part of Control Plan).
    AVY

    0
    #132370

    AVY
    Participant

    Gottu,
    The variation reduction should improve ur Cp and by the accuracy (shifting mean towards the target) should improve your Cpk. So there is no doubt that it would help.
    Whether shifting the mean will be expensive or not will be decided by the analysis done by your team and the improvement actions identified. When it comes to investment part to improve your process, you should do the cost-benefit analysis before taking decision to go ahead with the change. Ultimately the process experts always know, how the process can be improved and many times, you can refer the industry benchmarks to understand how much is the scoep for improvement.
    AVY

    0
    #132369

    AVY
    Participant

    Gottu,
    By reducing the 

    0
    #132362

    AVY
    Participant

    Gottu,
    Good quesions ? Whether the reduction in variation will result in reducing the mean cycle time will depend on whether the location of the spread.
    It always makes things easier to implement after you reduce the variation. Lets say, currently your mean is 88 min/pc. Now you may have different types of spread having the mean = 88. But the spread which is the narrowest, will have higher predictability and so you have better control on that process. On the contrary, if the spread is very high, your process is seldom within control and whatever actions you take maynot be consistent.
    By reducing the variation you improve your precision (consistency) and by moving the mean you improve your accuracy (nearer to ur target).
    Hope ur doubt is clear.
    AVY

    0
    #132357

    AVY
    Participant

    Hi Milo,
    I think, there is some clarity required in the way, you have phrased ur problem.
    I would like to know the following :
    1. What are trying to plot – parameter name ?
    2. What is the freq. of data collection ?
    3. Do you know the specification limits ?
    From what you have mentioned, you are using the X-bar and R-chart. Remember any point out of control limit need not warrant for the corective action. It can be treated as a special case and still discussed to move it to the favourable condition. If you are plotting the welding nuggent dimention, it may happen that lower spec. is more important than upper spec. ie. atleast the minimum nugget size has to be acheived. You can also use IMR chart which will help you to understand each dta point.
    AVY

    0
    #132356

    AVY
    Participant

    Onin,
    One  suggestion..
    If you want to proceed with attribute data, then just define your defect by setting the standards. eg. you have set 90 min/ pc as norm and anything exceeding that is a defect. In this case DPMO will be no. of instances (out of million), you have exceeded the target (doesnot consider by how much ?).
    If you want to proceed by using the variable data-which can be used to derive more information, you can mention the upper specification limit and then calculate the DPMO accordingly. (This takes into consideration the variation in your cycle time and you can understand where to focus on – reducing the variation or shifting the cycle time.) If variation is very high, you should focus on reduce tht first, so that ur process becomes predictable and then work on reducing the cycle time.
    All the Best !
    AVY

    0
    #132063

    AVY
    Participant

    Hi PGK,
    Human control cannot be referred to as ‘ mistake-proffing’. Mistake proofing is something, which will help you to prevent the defect to happen.
    In this case, depending on the severity of the failure mode, the process may have to be re-desgined, so that the process is ‘mistake -prooffed’. Eg. : We use swip card, which makes the system mistake prooffed, as no stranger can enter your office without using swip cards.
    In processes, wherein technology cannot be used, you should have a good mechanism by which atleast defect can be detected before it’s impact is seen at the customer end. For that you may have to do brainstorming, check with industry best practices, promote innovation etc.
    AVY

    0
    #132058

    AVY
    Participant

    Hi BB Boy,
    There are a few things that you may have to focus on immediately, before you decide to tak this as a improvement project.
    1. Get proper infrastructure in place – If you have a Union, you may have to sit and understand things from their point of view. Why would they participate in your project ? List down all possible reasons, for which they will participate.
    2. Discuss with your Dept. head and work out a proper recognition process, that will link the project savings to the recognition. Remember that recognition should be non-monetary, so that only the person who have received it would be benefitted.
    3. After the recognition process is worked out, you may have to roll out their improvement program, which will involve Six-Sigma project as a part of that.
    4. It might happen that some of the bright and experienced workers might come out with a brillant idea, (low lying fruit), which can give immediate results. Provide appreciation immediately.
    5. You may take up six-sigma project based on the scoping, complexity, time period and the cross-finctional involvement, which will help you to get to the savings target.
    All the Best !
    AVY

    0
    #131968

    AVY
    Participant

    Hi BB Boy,
    I agree with what Paul has suggested. The metric for maintenance is MTTR ( Mean time to Repair ie. – Avg. time to repair) and MTBF( Mean time between failures – Avg. time between failures). I will suggest you to do the following things for project selection :
    1). Pareto is used where 80-20 holds good .ie.- 20% of the Dept. contributing 80% of the problems. For 50% , pareto doesnot make any sense. Prepare pareto of downtime for equipment-wise for all the concerned Depts.
    2). See if 80-20 holds good and if yes, focus on the vital few from each Dept. If no, identify top 3 eqpmts. that contribute to downtime.
    3). Now, if you are already tracking MTTR and MTBF for each eqpmt., you may need to check the trend of MTTR and MTBF for these vital eqpmts. If you are new to this concept, start collecting data before it is too late.
    4). Define the problem for each eqpmt. and allocate improvement projects to the maintenance teams- responsible for maintenance of those specific eqpmt. ( Eg- Reduce MTTR from __ to __  or Improve MTBF from __ to __ .)
    5). This way you will be able to focus on all vital issues at the same time by addressing them at th esame time. You may need to understand whether the team is capable enough to solve the problem by themselves, else provide the mtraining ans support.
    6) Make use of FMEA for identifying the failure modes and taking corrective actions. If you alreayd have a FMEA, you may need to review it for any change in the process or any failure mode that is missed.
    Lastly, it is going to be a very interesting project and hope you enjoy this project. All the BEST !
    AVY

    0
    #131859

    AVY
    Participant

    Hi  Sue,
    What is th purpose of calculating the weighted Avg. of the yield. Can you enlighten us ?
    You may try to calculate the weighted yield by multiplying the daily yield% by the number that is the result of : scrap on that day divided by the total scrap generated during that pariod.
    Hope this will help.
    Regards
    AVY

    0
    #131817

    AVY
    Participant

    Hi Nandini
    You may have to get it clarified from ASQ itself, instead of getting an opinion poll. You may take the opinion, but it may turn out to be inaccurate and then you have only urself to blame.
    AVY

    0
    #131816

    AVY
    Participant

    Vinod,
    You may need to do the following taks, before you start with DOE.
    a.) Define the defect – What is considered as variation ? What is the measurement scale ?
    b.) Do Measurement gauge R&R ( I suspect this is an area for improvement )
    c.) Prepre detail process map and understand whether same process is followed by different operators ( skill levels or persons) for cutting hoses. Is ther any relation w.r.t the variation in the hose length and the operator.
    d.) You may few more factors after you do detail proces map and collect data on the potential factor. Based on data analysis, filter the significant ones.
    e) Only after you establish a relation between the factors and the hose length variation, think of including those factors in ur DOE exercise.
    Hope this helps. All the Best.
    AVY

    0
    #131815

    AVY
    Participant

    Fine ! What is your process background manufacturing or service industry ?
    AVY

    0
    #131812

    AVY
    Participant

    Hi
    The important thing, that ou may need to keep in mind is that all this done to understand the data (information) available with us to deduce something sensible our of this.
    Data stratification and sementation is most important in the Measure phase as this is the phase in which you prepare format in such a way that the information is gathered in proper mannre so that it will help you for further analysis.
    Subgroups and rational subgroups is done to clssify the data under similar heads to drill down the problem to the maximum extent.
    AVY
     

    0
    #131789

    AVY
    Participant

    Hi Amir,
    I am not from the textile industry, but I have a manufacturing background and I could tell you from my experience that usually the down time depends on the efficiency factor of the equipemnt used for production.
    I think, you may have to considered following factors, before you decide on the downtime figure :
    a). Equipment reliability factor based on the age and conditioning of the eqpmt.,
    b). The actual production capacity as compared to the design capacity ( you are under-utilising or over utilising the eqpmt., it can have an effect on the eqpmt. performance in the long term. Supplier should be able to help you in this)
    c). The life of the equipment ( depends on the age of the vehicle).
    d).The Quality standards of the product versus the equipment capability.
    e). Adequate Training of the technicians operating the equipment.
    The downtime should include any stoppage of production due to equipment problem or quality poblem or anything else.  
    Hope the above info. helps you in deciding the downtime.
    AVY

    0
    #64171

    AVY
    Participant

    Raj
    Thanks for the wishes and Hope the New Year will bring in Lot of Luc and fortune….the Lean way !
    AVY

    0
    #64169

    AVY
    Participant

    Hi Raj
    You are most welcome. You can contact me at my personal e-mail address – [email protected]
    AVY

    0
    #131720

    AVY
    Participant

    Hi Lijo,
    Donot worry, if you donot have anyone to copy, you can set an example by urself. Only thing, that is reqd. is having confidence in ur self and the team, in what u want to acheive.
    Back to ur question, can you identify the key customers and the deliverables in terms of cost, quality and delivery ? You may list them…….next step will follow after that !
    AVY

    0
    #64167

    AVY
    Participant

    Raj,
    The example of baby is not the right one to consider because , childbirth is not a repetitive process, that you do it in a daily manner. So, sorry mate, your example is not right.
    In simple words ‘ Lean thinking always prompts the person to questions, whether he/she is doing the right thing and also for the right reason ‘. It may happen that sometimes, you produce a product of very high quality standard, but your customer may expect it to be of not so high standard. In this case, you may be killing your process to achieve something, which neednot be done and  unnecessary inspection processes can be eliminated.  So waste can be anywhere – remember that delivery and qualiy are independent and there is no relation between them. Low delivery time (of any product) doesnot necessary mean that quality is better, so you cannot generalise anything between delivery and quality unless and until you have any data to prove that.
    Anyways, wish u all the best for ur lean journey. Are you aware, how to use VSM.
    AVY

    0
    #64164

    AVY
    Participant

    Hi Raj,
    You want to say that process highways are the process steps and the speed breakers are the NVA ( necesary and/ or unnecessary). In fact you have very well defined the CTD as the speed at which product & services are served to the customer.
    What remains mystery is the defination of quality from ur point of view ? How can quality get affected just because you try to drive the vehicle ( product/ services) at the right speed. In fact what matters is are you driving in the right vehicle ( right product / services are desinged or launched) and also are you driving on the proper process highway ( is process properly mapped that you can drive your vehicle as per the design).
    Example :
    If you have a SUV (Sports Utility Vehicle), you can drive on a bumpy road without any issues, but a passenger car willnot be confortable enough and your speed will get affected. So the most important thing is to drive on the highway, which is adequate for the vehicle to be driven. ( Processes to be designed in such a way that the desired services and products are delivered within the designated time period )
    Hope I am add some ‘value’ to our discussion.
    AVY

    0
    #131713

    AVY
    Participant

    J,
    IMR chart is used when you want to understand ‘ between’ variation i.e you are planning to capture TAT on a hourly basis, you can plot it, but you might see a pattern ( common cause variation) on a daily basis, which will hide the day to day variation as hourly data is poltted. So IMR chart is not desirable to use if you are trying to understand the day to day variation, but it will help you if you want to understand the hourly variation.
    Xbar andR chart will help you to understand both within (hourly) and between (daily) varition, so it is more appripriate to use the Xbar and R chart. Moreover the pattern will be take ncare of when you calculate the Xbar.
    Hope it helps you to take decision on what control chart to use and why to use ?
    AVY

    0
    #64162

    AVY
    Participant

    Hi Raj
    The value of a product or service can be both tangible as well as intangible. It is first important to define ‘ what do we understand by value ?’
    Value can be derived by using the CTC (Critical to Customer ) tree.
    It constitutes of :
    a. Critical to Quality – parameters that define the quality of product or service.
    b. Critical to Delivery – parameters that define the delivery aspect of product.
    c. Critical to Satisfaction – parameters that define the customer satisfaction.
    Preparing CTQ and CTD is not a tough task, but understanding the CTS needs lot of study, that connects you with the customer. CTS can be something to do with the feature provided to the product, which the customer might dream off or something that marks you different from your competitors- which makes you unique.
    So some of the parameters, that you may define under CTS may be very difficult to measure and those can only be known by doing a focussed survey or doing customer satisfaction survey / market research to understand where you stand in this.
    Value Stream Mapping :
    VSM can help you to define the value that is generated at each critical process step, in terms of quality, delivery and cost and it enables the team to understand the flow till the product reaches at the customer end.
    VSM is a very good Lean manufacturing tool, which help you to identify the excess fat ( waste / unnecessary processes), which does not add any value to the product. VSm works on the principle that, customer only pays for making the product and any process which doesnot contribute directly to the product value can be categorised into necessary waste or unnecessary waste. The key is to identify the unnecessary waste and save time and reduce te product cost.
    How is VSm related to value generation ? By defining the values of the product or service, you relate to each process step, in terms, how it contributes to these values defines through (CTQ,CTD and CTS). Any process that doesnot contribute to either of these needs to be eliminated.
    Hope it helps.
    AVY

    0
    #131649

    AVY
    Participant

    Hi Hanna,
    The purpose of control chart is to have a check on the process parameter by monitoring it within the boundaries of  +/-3sigma from the mean value. So it makes sense to use the control chart, when you are monitoring the actual value of process parameter rather then Sigma value. 
    If you want to understand within and between the day variations, it is recommended to use Xbar and R chart for your process paramter value. For calculating the sigma value, you should have atleast 30 data points, and if you are collecting data on a hourly basis, you can use Xbar and R chart for the day and then calculate the sigma value for the month. If you are operating in 3 shifts, you can plot the X-bar and R chart for shiftwise data.
    It is advisable to use just a simple trend chart, when you want to plot the sigma value for that day. Again, you have to understand, whether it is useful and what is interpreted out of that.
    AVY
     
     

    0
    #64160

    AVY
    Participant

    Gentleman,
    Your questions are not meant to be answered in this forum. You may have to ask your supervisor. If your supervisor is not able to help you, move to the next level.
    AVY

    0
    #131645

    AVY
    Participant

    Hi,
    You have asked questions which are very generic in nature ! You may have to start from -> the components of costs and then break it further into unit costs.
    ANother point, I would like to emphasize is that you may have to scope your project to a manageable level. Also, you have to have a measurment system for your ‘Y’. Can you define your problem in more detail, so that I can provide input properly.
    AVY

    0
    #131644

    AVY
    Participant

    Amol
    Takw my call ! ….. Donot look for readymade answers. You may have to develope a few things on your own and work on it further …
    Sorry, I cannot help you in this….unless you make an attempt from ur side to define what u feel it is ?
    AVY

    0
    #131643

    AVY
    Participant

    Hi  Neo,
    The Hypothesis test is based on the theory that ‘ Unless proven, the person is not guity of the charges ‘.
    The example quoted by you, has the Alternate hypothesis as Mean A not equal to mean B. So we have to look whether enough evidence is available to proven the alternate hypothesis.
     The p-value of 0.02 indicates that for Alpha value of 5%, enough evidence is available to say that the two means are significantly different. We donot check the alternate hypothesis, we only check the Null hypothesis. The inference drawn by you needs correction – There is 2% probability that the two means are same. i.e only 2% chances that null hypothesis holds true.
    AVY

    0
    #64159

    AVY
    Participant

    Hi Amol
    The Software product quality can be expressed as the multiplication of Quality, Cost and Delivery.
    All the above 3 parameters can be calculated for different phases of developement as mentioned below :
    a. Quality – It can have 2 measurables :
       i. Defect leakage rate and
       ii. Defect detection rate.
    Express the above parameters for per unit of development activity done Eg. per day or hour.Express it in %. Get a single figur by taking mutiplication of Defect leakage % and Defect detection %.
    b. Delivery – It can have following measurable :
       i. % variation planned vs actual dates
    c. Cost – It can have following measurable :
       i. % variation in the efforts (plan vs actual).
    Get a single figure for the Quality of software product by taking multiplication of –
    Quality rate X Deleivery variation % X Effort variation %.
    ( Note : The reason we have to get the variation because, we are looking at gauging the perdictability of the software developement process.)
    Hope you find it useful. Your views are welcome.
    AVY
     

    0
    #131637

    AVY
    Participant

    Hi Amol
    The Software product quality can be expressed as the multiplication of Quality, Cost and Delivery.
    All the above 3 parameters can be calculated for different phases of developement as mentioned below :
    a. Quality – It can have 2 measurables :
       i. Defect leakage rate and
       ii. Defect detection rate.
    Express the above parameters for per unit of development activity done Eg. per day or hour.Express it in %. Get a single figur by taking mutiplication of Defect leakage % and Defect detection %.
    b. Delivery – It can have following measurable :
       i. % variation planned vs actual dates
    c. Cost – It can have following measurable :
       i. % variation in the efforts (plan vs actual).
    Get a single figure for the Quality of software product by taking multiplication of –
    Quality rate X Deleivery variation % X Effort variation %.
    ( Note : The reason we have to get the variation because, we are looking at gauging the perdictability of the software developement process.)
    Hope you find it useful. Your views are welcome.
    AVY
     This thread has been moved to the IT/Software discussion forum. Please click here to continue the discussion.

    0
    #131636

    AVY
    Participant

    Hi Babu,
    The answer to your question is : Why not ?
    Since India is opned to Global market, we should be competitive enough to sustain the quality as well as continuously improve on the cost front. Six-Sigma is one of the initiatives, that can help the organization to get ther much faster. Since it is a proven methodology, there are lot of success stories that have been witnesses.
    You taking up the Green belt training has a lot to do with what is expected out of you from the organization, after the completion of the same. Hope you are aware of that and also able to take up projects after the training is completed.
    All the best
    AVY

    0
    #131635

    AVY
    Participant

    Hi Jim
    You may adopt the following process :
    Method A :
    a.) You may have to categorise the consultants under different  groups, with respect to the internal customer, that they are serving to.
    b.) The expectations from the internal customers needs to be brainstromed and prepare affinity digram for the same.
    c.) Next is charting down questions – which will help you to get correct feedback – in the sense that questions should not be generic but specific, so that the subjectiveness is minimised.
    Method B :
    a) I believe, you must have identified the key deliverbales for the consultants or may be SLAs from the internal customers.
    b) You may have to categorise under different headings and be more specific, to avoid any bias.
    Hope you find this useful.
    AVY

    0
    #131091

    AVY
    Participant

    Hi Buddy,
    You willnot get anything readymade, but you can gather the information in pieces and present it, in a way, that may make sense to your ‘yellow belts’. I will like to give you my input, as to waht all you can include as a part of that presentation :
    1. What is Six-Sigma ?
    2. Brief history of Six-Sigma and Organisations using it.
    3. Purpose of implementing Six-Sigma in ur Organisation.
    4. Six-Sigma deployment structure.
    5. How is the role of a yellow belt derived from the above structure ?
    6. Explain DMAIC cycle with 1 slide for each phase and you may show the outcome from each phase and the tools available to acheive those.
    7. Role and resp. of team members from the 6sigma project.
    8.Where is your organisation as per the deployment plan ?
    9. Share the benefits from the sic-sigma project, which they can relate to.
    10. Assign the ‘yellow belts’ tasks to be done after the overview session
    regards
    AVY

    0
    #64156

    AVY
    Participant

    The population is the collection of samples that can collectively provide information, which will help to infer about the population. So it is very important that whenever you collect the sample, it should represent the population.
    The Bell shaped curve for the population will have more spread because it is the collection of the samples over a period of time. The spread is more in the population due to the phenomenon of ‘ Shift and Drift’. The common causes will make the sample data to move to and fro and this will cause the population spread to increase for the period considered.
    Also, you will observe that for the sample data it willnot be normal because the datapoints are less compared to population. This is due to the ‘ Central limit theory’. Hope you are aware of this. You will get more info on ‘ shift & drfit’ and ‘ central limit theory’ in Statististics & Analysis under the Quality directory on the left side menu.
    Hope ur doubt is clear

    0
    #129441

    AVY
    Participant

    Hi Jenny
    What is CPA ?
    Usually for Six-Sigma projects, there will be separate role identified for calculating financial benefits, and that role is performed by ‘ Financial Analyst ‘.
    Typically, BB or the MBB should be able to break down the benefit calculation structure, whereas the Financial Analyst will take ownership for providing the benefit calculation sheet.
     

    0
    #64114

    AVY
    Participant

    Rakesh
    Explore the formula and explanation given in ‘ Capabilitiy Analysis ‘, available in the ‘Quality Directory ‘ on the left side screen if the isixsigma webpage.
    Hope it will be helpful.
     

    0
    #64110

    AVY
    Participant

    Hi
    Answer to ur questions is ” Yes, there will always be difference in the sigma value calculation of variable data and attribute data, for the same metric, because both are measured in different  ways. “
    In the example given by you – For attribute data, the defect is defined as ” Whether the downtime exceeds 60 minutes or not ?”. You will capture the results in ” Yes” or “No”. It doesnot lay emphasis on “By how much has the downtime exceeded 60 minutes”. It considers 61 minutes and 90 minutes at the same level. The DPMO will include the number of  datapoints exceeding 60 minutes of downtime out of 1 million datapoints.
    In the case of variable data, you will capture the actual value of downtime and it will treat 61 minutes and 90 inutes differently as this will get reflected in the ‘ standard deviation’ value. Greater the variation, more the Std. deviation and lesser the sigma value.
    To summarise, go for variable data whenever possible and use attribute data, only if there is no way to get the variable data.
    Hope your doubt is resolved.
    AVY
     
     
     
     
     
     

    0
    #128408

    AVY
    Participant

    Hi Johnny
    What u r missing by not using Six-Sigma methodology, is something I cannot comment unless I fully understand the way PPMI works, but I would like to mention some key points, which will help ur company to drive improvements more vigourously and improve bottomline using Six-Sigma, as listed below :
    a). Six-Sigma uses a systematic way of addressing any concern. The methodology ensures that there is no subjectivity involved in understanding about the problem by having a proper ‘ Operational Definition’. In traditional problem solving methods, we tend to overlook this factor and then after the improvement actiobs are taken, we release that the problem itself is properly understand by all.
    b).It links the improvement project to the top goal using the ‘ CTQ drilldown ‘. This way, ownership is driven to small teams and the teams know the improtance as it is linked to the top management goals.
    c) Many times, in the absence of proper ‘ metric’, we tend to drive improvement actions and lateron struggle to understand why the improvement is not visible. DMAIC methodlogy trains the people to use proper metric and in many cases question the ‘existing metric’. This helps the team think and do things that makes sense to them.
    d) After proper metric is identified, the measurmeent system has to be verified. In traditional problem solving methods, we tend to take the things for granted whereas in DMAIC we vrify the measuremen system, before we say ‘ how bad or good is our process metric’. This is a very big advantage of sic-sigma, which has a long term and deep  impact on the employees, in the way they interpret something as good or bad.
    e) In PPMI, there is no ‘ Control’. So after u drive the improvement actions, how do u sustain the benefits. In DMAIC, the control phase take care of this and this is the last phase before we say the problem is solved. So the problem is fully resolved only after the improvement actions are taken and the systems are put in place to sustain the benefits.
    f) In Six-Sigma extensive use of data is done, to conclude whether something is improving or deteriorating. So the perception based information is converted into hard data and this helps us to drive the improvement actions or justify investment for improvement purpose.            
    The above points are just to name a few. Hope the above points will help you to understand about sic-sigma better.
    regards
    Ajit 

    0
    #64076

    AVY
    Participant

    Hi Rob
    If your metric is defect leaked at each stage of developement process, you already have a variable data with you. So in the case of variable dat, you donot need the number of opportunities.
    If using the opportunties concept in Software develeopement, you need to identify all the falure modes and for that requirement capturing process has to be very robust. When you prepare the unit test plans, you also identify the expected output. The number of expected functional reqmnts. can be taken as the opportunities in this case.
    Hope this helps.
    regards
    AVY
     

    0
    #64067

    AVY
    Participant

    Hi
    This is a very interesting situation and is very commonly encountered whenever you reduce the cycle time.
    The anwser to your question, whether your process is anti-lean or not is “Yes”, if you are have inventory excess of your plan.
    What to do next ?  Remember that, you can only control your inputs and not your customer orders. End of the day, you need to optimise the process to get maximum revenue.
    You have to first understand the inventory cost of 1 unit of excess inventory. The point at which  the inventory carrying cost is more than the target or maximum limit, that should be the breakeven point for you to stop  building any extra units. Instead of building no.of units using theavailable resources, you may have to use resources only based on the units to be build. So any of resources that have resulted due to improvement in your process can be used for some other process or activity instead of keeping it ideal or engaging them inproducing excess goods. Remember you can only eat sufficient quantity, that you can digest……
    Lastly you may have to consider lot of factors before you allocate your resources. Probably the resources that need some training can be trained to improve or skill upgradation can be done, which can help your team in the long term and also improve their morale.
    regards
    AVY

    0
    #126395

    AVY
    Participant

    Six Sigma dude
    Project selection process is a very Basic question that u are asking. If you have doubts at such a basic level, then probably you need to undergo the Black belt training once again.
    Though there will be two opinions on who identifies the problem, it is very impt. that the Process owner agreement is done before the BB decide to take it as a project.
    You may have to understand what is suitable for your organization and also what will work in the lon term.
    AVY

    0
    #126391

    AVY
    Participant

    Hi Ramesh
    You have asked a very good question and many BBs have this question in mind, when they start with their first Six-Sigma project.
    As per my opinion, the first question, you need to ask yourself is          ” What are the factors affecting your Big Y “. You may use correlation analysis to filter out the factors and check if there is a linear relation between the input factor and the output. Many times, there is interaction between the factors and the significance of the factor is only highlighted when 2 or more factors come together. Now,  for this you can use Multi-vari chart, which will help you to understand the interaction between the factors and their effect on the output.
    After you identify that these 2 or more factors are significant, next step will be to use regression analysis to find out what is the significance level of this interaction. This will help you to understand how much % of variation in the Y can be tackled by addressing these 2 or more  factors.
    After regression analysis, you will have a confidence level that will tell you much of variation is answered by the regression model generated out of the data available. How do you know, what is the optimum level of these factors that will give you the desired output ? For this you may use the DOE tool to optimise ur X’s.
    The Best way to understand the usage of these tools is to identify what kind of analysis is required to be done at any particular time of the project phase.
    Hope this helps you. Anyway all the best !
    AVY

    0
    #64046

    AVY
    Participant

    Rob
    You need to have a proper Operational Defination of ” Change Request”.
    You need to focus more on the kind of “Change requests”.- like whether it is related to cosmetic change or functional change which was identified in the requirement phase. The purpose for doing this analysis has to be understand first.
    If your data is not normal, convert into normal using Box-Cox transformation and then do the statistical analysis.
    AVY

    0
    #126040

    AVY
    Participant

    shook1
    That is a good initiative. In my opinion, the success of this initiative should have metrics under 3 categories – Cost, Quality & Delivery. You need to understand how you are performing in the current set-up of onsite for these metrics and then have some targets for the offshore, that is sensible enough.
    Cost :
    – Cost of detecting one defect : This is the function of no. of defects that are detected and the time taken to detect those defects.
    Quality:
    – Rate of defect detection : When you move your onsite testing to offshore, your rate of defect detection should not get reduced. If it is improved , it means that either test plans are improved or the testing team is more capable. You can provide incentive to motivate the offshore team.
    Delivery :
    – Delivery of testing in-time. If your offshore team is efficient than in a given time, they can test more than the onsite team. This can justify the offshore activity also.
    Hope you found it useful.
    AVY.

    0
    #126039

    AVY
    Participant

    Mukesh
    Waht do u actually need ? How to present your GB project or What GB project you would like to select ?
    PL. mention waht support you need from us.
    AVY

    0
    #125983

    AVY
    Participant

    Hi
    First thing, you need to do is to understand the consept of ‘ Process Capability’. You will find lot of information at the isixsigma  website under the Quality Directory. Click on the Statistics and Analysis and you will get your question answered.
    To cut everything in short, you need to understand the following things before you quantify the problem :
    – Frequency of data collection should be on a daily basis. If you have daily data, then it makes sense to plot the control chart.
    – In case you plot control chart for the defect data, check for normality and then calculate the capability for the customer specified USL & LSL.
    – Once you calculate the capability, you can get the current sigma level for the defect. Putting in Sigma terms, helps us to understand both the Mean and Std. Dev. for the data collected.
    I would recommend you to collect the defect data on a daily basis and then quantify the problem. Collecting data on a monthly basis is not appropriate as you may not remember what went wrong 2-3 weeks back. So daily tracking is recommended.
    AVY

    0
    #64039

    AVY
    Participant

    Hi Shamsi
    How can FMEA be used in the Requirement Developement phase ? Accordingly to me the requirement developement phase will include the Voice of the Customer ie. the expectation from the customer end, what the software is supposed to deliver ? How can FMEA be used here ? 
    I can understand if it is used in the Build phase  (developement phase), wherein we try to understand what can go wrong in the way we are developing  the software. Failure modes can be identiifed and priorities can be decided, so that proactively actions are taken.
    Your views !! Do you have any example, wherein you can share what you are saying ?
    AVY

    0
    #64034

    AVY
    Participant

    HI BBC
    In fact, many people are unaware that till a BB is certified hei just a ‘BBC’ ie. Black Belt Candidate.
    I have used a wide variety of Six-Sigma tools in IT  projects. Let me now, if you need any help in this. Send me ur e-mail address.
    AVY

    0
    #125201

    AVY
    Participant

    Hi CT
    The name ‘ Six-Sigma ‘ is just used for Branding. Similarly, the BBs and GB s are nothing but the titles that will give credibility to the people who ar involved in improving the bottomline.
    The BBs undergo a training for 4 weeks that spans across 4 months and the GBs undergo a 3-5 day program. The type of projects that a BB undertake, have a wide scope involving cross-functional teams and use of advanced statistical tools. For GBs, baisc statistical tools are taught and over a period of time, as they work closely with the BBs, their knowledge lev lcan incease, but it may still not fulfill the BB requirement. Again, you will have lot variability in how the GBs or BBs are trained.
    Again getting GB certification will definately help the person in understanding Day to Day issues and using appropriate Six Sigma tools.
    The most important thing that differentiates a BB from Gb is that BB has to have strong leadership qualities whereas a GB neednot have  strong leadership qualities. Also in a matured organization, you will find that 20% are BB s and 80% are GBs.
    AVY.

    0
    #125013

    AVY
    Participant

    Joe
    All the statistical formulase have been derived based on the assumption that the data is normal. That only means that the data is predictable to the extent that it the histogram will follow a bell-shaped curve.
    Statististics is a study of  samples representative of the population and infering about the population. If the sampel size is not good enough (minimum 30 for variable data ) or if the sample is not representative of the population ( biased data collected like only data from shift-A only etc)or of the data is biased towards one side of the specification i.e on-sdied spec., you may have to convert it to normal and apply statistical formulae to the converted values.
    Unltimately, the probability that your inference about the population is good enough will only hold good, if the data is predictable and followign a certain trend.
    AVY.

    0
    #124928

    AVY
    Participant

    Srinivas
    You have the information almost available with you. As you read this, click the ‘ Methodologies’ under ‘ Quality Directory’ on the left side of the screen.
    AVY

    0
    #124793

    AVY
    Participant

    There have been lot of thought process put in by all the respondents. I would like to also add to it ..2 methods are proposed…
    Method A :
    1. Prepare a detailed process map.
    2. Prepare a ‘ Affinity Diagram ‘ for the 90X’s. Affinity diagram will help you to bring together the X’s that are related to each other.
    3. Assign the X’s to each process step and filter the 90X’s using FMEA methodology.
    Method B :
    Using Cause and Effect Matrix, prioritise the 90 X’s. In the Cause and Effect matrix, you will have to first list all the process Y’s (it can be one or more) at the top row and rate that Y on the scale of 1-10 importance to the customer. Then list the 90 X’s on the left side in the column. Now rate each X on the basis of effect on that particular Y. Get the final score by multiplying the score for each X and adding the 2 score for the 2 Ys for each X. Priortise, the X’s based on the score.
    This is a tried and tested tool used very effectively, when you have to filter the X’s down and lot of expertise is available to rate the Xs.
    It is up to use to choose the right tool. Untimately timeline is one thing that need to taken in consideration, when chossing the right tool.
    AVY

    0
    #124690

    AVY
    Participant

    Hi
    OEE stands for Overall Equipment Effectiveness.
    Let me give you a simple example to make you understand how this metric is used.
    OEE is a metric typically used in a manufacturing  industry. Let us take an example of a m/c shop, consisting of 10 machines. Now, as a plant manager, you should be in a position to understand – whether you are using this equipments to its planned capacity. 
    For that you have to calculate the OEE for each of the equipment and check up whether all the equipments are delivering production and quality at the expected performance.
    AVY

    0
    #124597

    AVY
    Participant

    Paul,
    I have never said that what u r following is wrong, instead I was just trying to explain Vidya, the OEE concept at an higher level.
    What u r following is a very good system, which might work very good for ur organization. I just shared what was being followed in my previous organization. Afterall, we r here to share information.
    Hope, Vidya has got what she had asked for !
    AVY

    0
    #124593

    AVY
    Participant

    Hi Balan
    Your First Question : Is there any statistical method, I can use instead of increase the spec. ?

    I am assuming that in this case your  measure is attribute data. ie. Leak test  – Ok / Not Ok.
    Check the measurement system – Do Gauge R&R . Understand the variability due to the leak test eqpmt., operator and the method followed.
    If gauge R&R is not OK, take appropriate actions and redo the MSA, till the gauge R&R is passed. When investigating, try to understand whether something has varied over last few weeks or days like the operator or the eqpmt. breakdown- that might affect the leak test.
    Is your MSA result is not OK, you may have to check all the parts for the leak test and probably also check some of the rejected parts – r they really defective.
    If the MSA is Ok , checkup daily trend of parts rejected.
    Also, prepare a detail process map and do FMEA analysis of what might have gone wrong with the leakage. ( You must have alread yhad that). Collect data on the input parameters and checkup if any i/p parameter is out of spec. or have changed significantly over a period of time. Use approrpaite hypothesis test, if applicable.
    Your second Questions : I want to test the product using the original spec. Vs increase spec. Is that method can tell us something ?

    You can check the same parts using the same method, equipment and operator, to check how many parts pass the leak test.
    I donot think, the above comparision can indicate anything, other than the prediction of no. of  parts that you will fail to reject because of increase in spec.
    Anyways, why do the management want to increase the spec. ? IS it going to fulfill the functional reqmt. of the product ? If  yes, then will the life of the product get affected ? If yes, then go ahead and supply few parts with your customer’s consent and implement it only after the customer gives a go head with the change.
    AVY

    0
    #124591

    AVY
    Participant

    That was a very well prepared sheet. Vidya can now prepare a similar sheet with formulas.
    Just one input, I would like to add further to what Paul has provided. Under the Availability heading the S1,S5 & S6- it is recommended to capture this under the ” Tool Change-over Time “. If it is a planned repetitive  activity, it is advisable to track that time separately. This will help to get the true picture of your process and also since the Tool change-over time is a tracked separately, you can monitor it separately.
    The OEE should typically tell you given the available resources (availability) how are we performing in delivery (Performance efficiency) and Quality (FTT). Keeping this in mind, the OEE calculation methodology can be finalised.
    AVY

    0
    #124460

    AVY
    Participant

    Naren
    Remember ! you can only control what you can measure, at the same time, it might happen that, many times what and how we measure, should make sense to the process owner.
    In this typical case, it is the employees, who are the process owners. According to me, it is a good idea to express the performance in Sigma value – for the Dept. As for the Dept, you have clear cut objectives that have to be met. Wheras for the employees there should be some form of checklist…list of objectives for the year and how many have been met. The percentage of objectives that a person has met or exceeded can be used as a metric, to decide whether an employee has overperformed or underperformed. This can keep the things very simple and understandable.
    Am important question you need to ask before you implement this is – can the process owner track the metric himself. In this case, if the employees have reached to a maturity level, wherin, they can calculate the sigma value themselves, then go ahead and implement it , else be cautious. Remember the fundamentals – never try to take the ownership from the process owner, else the system will collapase.
    So my advise is : – First calculate the sigma value for the depts. Then implement the checksheet as proposed in the above paragraph and then let the employee take charge of their objectives. Leave them alone. They will themselves come up with Sigma calculation concept, if it makes sense to them. Either way, the performance is tracked and ownership is driven to the bottommost level.
    AVY.

    0
    #124458

    AVY
    Participant

    Shirshen
    First of all, an unstable process need not be necessarily be a combination of multiple processes. Why ? Read below ..
    An unstable process is only an indicator, that you should be Looking into. Let us take an example, when an unstable process consists of only one process step. In  that case, there will be no other process step you need to focus on ! Also, before you start working on a process input, there should be enough evidence with you that the variation in that i/p parameter is causing the variation in the output. Just because, there is variation, doesnot mean that that input parameter is a problem. It might happen that there is some that noise factor that can cause the output to vary and that might b missed out.
    So, be cautious about relating the improvement to variation in i/p. Only after enough evidence of relation is established, only then that i/p variation is the culprit.
    AVY

    0
    #124449

    AVY
    Participant

    MKS
    You have not mentioned the scope of your SS project. Ideal way to begin the project is by preparing a high level process map and understanding the processes that you will focus on, for the SS project.
    I would liek to know, in what phase, ur project is currently. Is it in the improve phase or define phase. Pl. clarify ? If in the omprove phase, what are the rootcauses identified for improvment, so that we can provide suggestions accordingly.
    AVY

    0
    #64032

    AVY
    Participant

    Hi TamiSz
    First of all, let us understand the Basics of statistics.Ther are basically 2 types of ;imits we refer to in statistics – the specificatio limits and the control limits. Now what are they ? let us see…
    The Specification limits (USL & LSL) is the voice of the customer. It indicates within what range the customer expects the product parameter to vary. This will always remain constant, till any change is made in the SLA(Service Level Agreement) or the control plan or contract.
    The Control limits (UCL & LCL) that are shown in any control chart, is the voice of the process. As your process parameter changes over a period of time, the same will get reflected in the control limits. Pl. note that the control limit is ‘expected to vary’ and not remain constant. If the control limits remains constant, it only means that your measurment system is not good enough to measure the variation. The UCL and LCL are calculated by using the formula : X-bar +/- 3Std. Dev. Since for every sample data, the mean and the std. deviation can change it is not expected for the control limits to remain the same.
    The next question, that you may ask is – why is then the control limit required ? The control limit indicates how is the process performing over a period of time. Also it is always advisable to plot the control chart manually, instead of using minitab. When you plot the control chart for the first time, you have to check the normality and then, calculate the UCL & LCL, only if the data is normal. After that for the data sample 2 – use the UCL & LCL from the 1st chart and then plot the chart on a real time basis. So whenever there is any point going our of control, you have to go back and immediately check ur process and take appropraite action. After the 2nd chart is completed, recalculate the UCL & LCL and use this for plotting the 3 rd chart.
    So as you move on from 1st chart to 2nd chart, you are able to have a check on how the process is performing over time. It might happen that the control limits are within the specification limits, but it may still vary to the extent that you might have to ask questions, that will prompt the process owner to take actions to reduce the variation.
    Hope, you are able to understand the concept.
    regards 
    AVY                                                                                               

    0
    #64028

    AVY
    Participant

    Darin
    Typically IT processes are always client based. So you may not find anyone, who will be eager to share the benefits with all of us. Also, IT industry is yet to reach the maturity level, wherein the organizations start sharing their business measure for overall improvement in the benchmark. Currently there are no standard metrics available, that is used across all IT industries. Every IT company uses one that is suitable to their business.
    So donot waste time in waiting to get answers to this question. Instead go ahead and under your organizations delieverables and link it to ur IT dept and check for any gaps for improvement.
    AVY

    0
    #64027

    AVY
    Participant

    Daniel
    Before making the FMEA, you need to prepare the detail process mapping of how the IT process is used in Bank ie specifically in your Bank. After that, you need to identify potential failure modes for each process step by listing the process steps from the process mapping to the FMEA format.
    Most important thing that you need to keep in mind is that FMEA is prepared by the team members who contribute to the process and not by any one who knows about the process. This is because only the process owners should be in a position to give the severity, occurance and the detection for each failure mode. So I would recommend you not to ask any outsider for any suggestions for ur process. Follow the methodology and you will get to the rootcauses.
    Remember your role as a BB or GB is of a facilitator and not as a process owner. Go ahead and start preparing the FMEA…
    regards
    AVY.

    0
    #63999

    AVY
    Participant

    Hi Naseem
    First and the foremost thing to remember is that you need to be trained in FMEA, in order to prepare FMEA. If you are doing FMEA for the first time, you need to get an experts’ help to start with. Rather than the tool, the most important thing to remember is – the facilitation skill of the BB. It involves the process owner and the process expert and the BB needs to facilitate the FMEA meeting.
    The steps for proceeding for FMEA are :
    – First prepare a detail process mapping, wherin you identify all the process steps.In ur case prepare the detail process mapping for the migration process.
    – List all these steps on the left side of the FMEA template and then for each of these steps identify the critical parameter i.e. For each step of the migration process, identify the critical parameters.
    – For each critical parameter,you need to identify the failure mode. What can go wrong in the parameter ?
    – Then you have to follow the FMEA process….Hope u r aware of that.
    Since we cannot use this forum for training purpose, I will stop here. You may get in touch for any help .
    regards
    AVY.

    0
    #63997

    AVY
    Participant

    Sanjay
    There are 3 main criterias that need to considered when differentiating between a Black Belt & Green Belt project. They are as follows :
    1. The duration of the project : A BB project spans between 6-8 months, whereas GB project spans 3-4 months.
    2. The scope of the project : A BB project may need a cross-functional team to focus on, whereas a GB project may have team members from the same Dept.
    3. The Tools used for the project : Typically a BB project may require use of advanced statistical tool like Regression, DOE etc, whereas a GB may not need basic statistical tools.
    Also, it goes without saying that a BB project should provide significantly higher savings as compared to GB projects. At least a BB project should be addressing a more complex problem than that of a GB project.
    regards
    AVY

    0
    #121535

    AVY
    Participant

    Hi Franz
    Yours is a typical case when Six Sigma is implemented without buy-in from the process owners. Some of the key questions which you should ask before deciding the BB certification are given below :
    1. Were the BB projects customer focused ?
    2. Were the BB projects focussed towards the Dept goals ?
    3.  Is Six Sigma project benefit sustained after the project is closed ?
    4. Are the process owners actively involved in implementing the improve and control actions ?
    5. Have  the Six-Sigma projects helped ur organization to improve the key business goals or the bottomline ?
    If the answer for all the above questions is Yes, you have a good reason to recognize the efforts put by the Six Sigma team, else you need to redeploy Six Sigma. You may have to understand what has gone wrong in the implementation phase. If any one or more of the above questions have a negative answer, you need to think and change the way of doing things so that Six Sigma is successful.
    You need to remember that BB certification is only a recognition of the BBC’s skill to apply Six Sigma tools to improve the bottomline. If this intent is not met, what is the use of BB certification ?
    bye
    AVY 
     

    0
Viewing 64 posts - 1 through 64 (of 64 total)