Fatigue Calculation?
Six Sigma – iSixSigma › Forums › General Forums › Tools & Templates › Fatigue Calculation?
 This topic has 1 reply, 2 voices, and was last updated 5 years, 10 months ago by Robert Butler.

AuthorPosts

March 7, 2016 at 4:54 pm #55274
68rs327Participant@68rs327 Include @68rs327 in your post and this person will
be notified via email.Is there such a thing as a Fatigue Calculation or Calculator?
Example: I have an assembly line that assembles and lifts parts that weigh 33 pounds each. Though out the day, the cycle time of the line slows, thus output slows.
People are manually loading the parts into the machine, machine cycles, operators lift parts to next machine, machine cycles, and so on. This condition happens approx. 10 times per part. When I cycle time the line in the morning when people are starting work feeling rested, cycle time is fast and output is better. As the day goes on and people get tired from lifting the heavy parts, cycle time increases, thus output decreases.
Question: Is there a calculation that I can use to calculate a “Fatigue Factor”? We calculate OEE of the line using the bottle neck cycle time. But this cycle time was captured at the start of the shift. Should I take an average of the cycle time at the start of the shift VS. the cycle time at the end of the shift to determine a more realistic OEE?0March 8, 2016 at 5:37 am #199422
Robert ButlerParticipant@rbutler Include @rbutler in your post and this person will
be notified via email.Based on your description it would seem that the best way to characterize your process would be to build a regression model with cycle time as the outcome and measurement time as the predictor. That would give you cycle time as a function of elapsed time along with an estimate of the variation you could expect to observe at any time during the day.
The big issue is that of data structure. Your post gives the impression that you only take cycle time measurements at the start and the end of a shift. If true you would want to increase your sampling to something like once every two hours for a given period of time in order to have sufficient data for model building. You might also have to think about the issue of repeated measures.
As I understand your current system of measurement you are measuring the time it takes a given operator to complete one part and that time measurement is the amount of time it takes that operator to complete a 10 step process. If this is the case then the smallest unit of independent measures is a given operator. This means that the cycletocycle time measurements for that operator are not independent events. If the above is true then you will have to use repeated measures regression methods to accurately model cycle time as a function of time of day. Since you mentioned work shifts you would also want to include a dummy variable for shift to make sure that there are no shifttoshift differences.
0 
AuthorPosts
You must be logged in to reply to this topic.