iSixSigma

What type of statistical distribution is this data?

Six Sigma – iSixSigma Forums Old Forums General What type of statistical distribution is this data?

Viewing 7 posts - 1 through 7 (of 7 total)
  • Author
    Posts
  • #41614

    Dwib
    Participant

    Hi,
    I am looking at a project to reduce queue time with our truck fleet. The perfect situation is for a truck to arrive as the previous truck leaves. Therefore the aim is zero queue time. But this isn’t always possible. Therefore I would assume if you plotted a distribution graph you would see the majority of the data clustered to the left and the tail trailing out to the right.
    My understanding is that this is not a Normal Distribution. If so then what type of distribution would classify this type of data?
    The reason I want this information is so that I can see how the data fits to the correct distribution in Minitab.
    Thanks in advance.
     

    0
    #130657

    Carl
    Participant

    Dwib,
    This sounds like it would be a Weibull distribution.
    Good luck,
    Carl

    0
    #130661

    BTDT
    Participant

    Dwib:If all truck arrivals are independent, then the interarrival time will have an exponential distribution with a mean calculated as (number of trucks)/(time interval).The exponential is a special case of a Weibull distribution with a shape function of unity.If you have trucks waiting for a vacant bay, then the distribution will be a shifted Weibull/exponential.There are equations for predicting the average time in the queue and length of time in the queue given some general assumptions about the service time, the number of servers, etc. Have a look for them and use an exponential time for interarrivals and you should get a leg up on your capacity planning.Cheers, BTDT

    0
    #130907

    Todd Hatley
    Member

    Much of the literature classifies the arrival rate as a Poisson process.  There are lots of interesting things that you can analyze related to this problem utilizing queuing theory. 
    I am out of town but will try to post some reference when I return home.
    —Todd

    0
    #130917

    Anirvan Sen
    Participant

    The data distribution is skewed to the right. Depends on what do you want to do with it. Give some more details.
    Anirvan

    0
    #131034

    Dwib
    Participant

    Thanks to all that have replied.
    I would like to reduce the amount of queue time on average for our truck fleet. Therefore I wanted the correct statistical distribution to calculate the process capability of this process.
    My thoughts are to:
    1) identify the current process capability and use this as our baseline.
    2) Analyse the data we have with this process to identify particular areas we can focus our efforts upon.
    3) Present the information to our team to identify some prospective solutions.
    4) Implement the solutions
    5) Recalculate the process capability once the solutions are embedded into the new process to see if any improvement has been made.
    Any comments on my proposed steps or supporting information on this topic would be appreciated.
    Regards
     

    0
    #131038

    jimmie65
    Participant

    3.5) Test the proposed solutions before full-scale implementation.

    0
Viewing 7 posts - 1 through 7 (of 7 total)

The forum ‘General’ is closed to new topics and replies.