BITS Pilani Goa campus is having its annual technical fest 'Quark' , so we decide to conduct a social experiment to check our competency in acting smarter than we actually are to people we should'nt even be trying it with. It was fun and we've valued some important lessons along the way like ,
1 - ALWAYS BE ON TIME: We reached in the early AM at the venue for the six sigma workshop later that morning. How did this help us ? , we thankfully got a room for us even though we werent 'officially' registered for accomodation, ahead of the hordes of people who kept pouring in throughout the day in search of shelter.
So KPMG was conducting a 'Yellow Belt' Certification program through the day , which strictly speaking , is an infantiile initiation into the Six Sigma family.
Whats Six sigma you ask ?
Six Sigma is the confidence or gaurantee of output at 99.99967% to CTQ parameters
99.99967%
But isnt something we all know like 99% enough of a target to achieve ?
Is it really required that this marginal improvement be made ?
Heres an example of failures occuring at 99%
400 articles of lost mail per hour
3000 newborns falling from hands of nurses
And failures occurring at Six sigma
1 lost article of mail per hour
10 newborns falling from hands of nurses
What Six sigma aims to achieve is the reduction of deviation from the mean.
So we essentially need to reduce process variation/inconsistency.
Variation is statistically defined as sigma , in this case we aim to control ( +/- 6 Sigma ) from the mean.
So what we are trying to say here is Y=f(X)
Y= output
X= input
As long as your output varies as a function of the input. So take care of the inputs (and functions varying it), that directly relate to CTQ (critical to quality) parameters.
In order to have Six Sigma capability, the variation would be small enough for you to fit plus or minus six standard deviations within customer expectations.
Thats fantastic ! , so how do I achieve this transformation
With the DMAIC approach
DEFINE:Zero in on a specific problem with defined return on effort
MEASURE:Determine current process performance and identify improvement measures (basically a lot of statistical data like paretos,fishbones etc)
ANALYSE:Validate the process for improving performance.(potential cause is actual or not ?)
IMPROVE:Improve performance and validate realized results.(brainstorm/psychological inertia/refinement stage)
CONTROL:Implement controls to ensure continued performance(revisit control charts)
thats it in a nutshell.
2 - THERES ALWAYS ROOM FOR ONE MORE ! Upon our return to our hostel room we were introduced to this maxim ,when it appeared to us that the organisers had recurrant problems with the still approaching hordes of visiting crowd , so we were 4 adults in a room built with enough air for one ; whence my comrade spake his oft quoted words.
We hit the next days sunrise with a brisk walk to the nearest bus stand , heading to Panjim, onwards to Calangute - Baga ; back to panjim , with the following customary stops.
Miramar beach , souza beack shack , various road stalls, Brittos , beach nap , long bus journey, miramar beach .
But of course the highlight was the part where we were desperately trying to flag down vehicles en route to BITS on the deserted stretch of airport road at 10 PM ( local equivalent of 1:45 AM ).
3 - IF YOU SHOULD SEE ME ON A DARK ROAD AT NIGHT , PLEASE PULL OVER !!
