iSixSigma

Choosing Samples for VSM

Six Sigma – iSixSigma Forums General Forums Methodology Choosing Samples for VSM

Viewing 6 posts - 1 through 6 (of 6 total)
  • Author
    Posts
  • #55576

    John Pelentrides
    Participant

    Hi There,

    I am about to develop a VSM for one of our cells in the factory and am having difficulty choosing samples to define each of the processes within the cell

    There are around 150 different products that flow through the cell all with different cycle times, up-times, etc.

    I was considering choosing around 15 samples but am struggling to decide which products to choose. I was thinking of using the most common parts going through the process but the might not give a good representation of causation of issues.

    Any advice would be greatly appreciated!

    0
    #200584

    Andrew Parr
    Participant

    @jpelentrides

    I can but guess. What is the problem you want to solve?

    0
    #200585

    John Pelentrides
    Participant

    @Andy-Parr

    I have asked to find areas in the manufacturing process to improve. This cell produced this lowest yield % (highest scrap rate) and therefore chose this cell as a good place to start.

    I want to develop a VSM to analyse the cell at the top level to highlight the best areas to start focusing time on.

    The only issue is a amount of different products that pass through the cell. Due to there being such a high number, it wouldn’t be an efficient use of my time to analyse every product.

    I was thinking of attempting to rate the products using volume and value. Some advice on choosing which products to sample would be ideal.

    0
    #200591

    Andrew Parr
    Participant

    So how much data exists on the scrap? Doesn’t that tell you where to look?

    15 samples of 150 products is playing Russian Roulette if you expect to hit on the problem so you need some clues from somewhere. It looks like the scrap rate data should give you the comfort of where to look for most reward.

    If you know the cell has the highest scrap rate then you need to see where that scrap comes from. You also might ask if all other cells produce the same products. You’d need to analyse but changing from one product to another might be a source of defect.

    Just my opinion but 15 samples is too few by the look of things.

    0
    #200596

    MBBinWI
    Participant

    @jpelentrides – are there common types of defects/scrap? Instead of looking at parts/products, I would look at types of defects being created.

    0
    #200600

    John Pelentrides
    Participant

    Thanks for your feedback @Andy-Parr and @MBBinWI. It is very useful to hear other opinions as I am new to this.

    To put this into a bit more context:
    This is a finishing cell for a casting process. This cell looks into correcting inspecting individual parts before being sent either to machining or to the customer. Within the cell there are multiple operations including mount point, straightening and inspection.

    There is already an in depth investigation into defects parts being carried out being lead by myself and a process engineer.

    This project is looking more into movement of parts and operators, time parts are waiting for each operation and time taken to complete each operation.

    I was planning of developing VSMs (collecting process data and inventory in supermarkets) and spaghetti diagrams for each cell in the factory one-by-one as there is no current state VSM for any cell in the factory. My issue is that there is so much data to sort through it would be an inefficient use of time to analyse it all. I was trying to work out the most accurate way of analysing each cell without sorting through millions of bits of data

    0
Viewing 6 posts - 1 through 6 (of 6 total)

You must be logged in to reply to this topic.