Six Sigma in Statistica 

What is „Six Sig­ma“? 
Six Sig­ma is a struc­tu­red, data-ori­en­ted method to redu­ce defects, was­te and qua­li­ty issues of any kind in pro­duc­tion, ser­vices, manage­ment and fur­ther busi­ness acti­vi­ties.  Six Sig­ma is a trade­mark of Moto­ro­la, that shaped the term in the eight­ies. The method is based on a com­bi­na­ti­on of estab­lished tech­ni­ques of qua­li­ty assu­rance, basic and advan­ced methods of data ana­ly­sis and sys­te­ma­tic trai­ning of the staff invol­ved with the Six Sig­ma rela­ted pro­ces­ses. 

Why is Six Sig­ma popu­lar? 
Six Sig­ma has shown that it not only increa­ses the qua­li­ty, but also that it can lead to tre­men­dous cost reduc­tions. Some spec­ta­cu­lar suc­cess sto­ries from huge com­pa­nies have been published. Jack Welch (the for­mer CEO of Gene­ral Elec­tric, one of the world’s lar­gest manu­fac­tu­ring com­pa­nies) has said: “Six Sig­ma is the most important initia­ti­ve GE has ever undertaken…it is part of the gene­tic code of our future lea­der­ship.” Welch attri­bu­tes Six Sig­ma for bil­li­ons of dol­lars in cost reduc­tions.  
Many other com­pa­nies cla­im huge cost reduc­tions after intro­du­cing Six Sig­ma at their pro­duc­tion sites, as well. Moto­ro­la (one of the lea­ding mem­bers of the group of com­pa­nies that deve­lo­ped the Six-Sig­ma-approach) for exam­p­le claims to have saved 11 bil­li­on dol­lars after intro­du­cing Six Sig­ma twel­ve years ago. Allied Signals men­ti­ons cost reduc­tions of about a bil­li­on dol­lars over a cou­ple of years. 

Sta­tis­ti­cal Back­ground 
The term six sig­ma repres­ents the sta­tis­ti­cal goal to redu­ce the num­ber of defects to a negli­gi­ble amount, cor­re­spon­ding to the six sig­ma value of a (cor­rec­ted; see below) nor­mal dis­tri­bu­ti­on cur­ve: Six Sig­ma tri­es to push fail­ures and qua­li­ty issues to the outer bounds of the dis­tri­bu­ti­on to redu­ce the pro­blems to rare out­liers of an other­wi­se fault­less pro­cess. 
To reach this “Six-Sig­ma-Goal” a pro­cess can­not have more than 3.4 defects per 1 mil­li­on pos­si­bi­li­ties, whe­re defect is defi­ned as any form of unde­si­red out­co­me for the pro­cess under inspec­tion. Plea­se be awa­re that 3.4 fail­ures per mil­li­on in fact repres­ents a z-value of 4.5 of the nor­mal dis­tri­bu­ti­on ins­tead of 6 as the method allows for a dyna­mic shift of 1.5 sig­ma (defi­ned by Moto­ro­la as long term dyna­mic mean varia­ti­on). 
The­r­e­for one of the basic Six Sig­ma rela­ted tools is the Six-Sig­ma-Cal­cu­la­tor to cal­cu­la­te the num­ber of fail­ures for one, two,  .., six sig­ma Pro­ces­ses. Natu­ral­ly, the­re are much more advan­ced tech­ni­ques to be appli­ed based on the pro­ces­ses over the dif­fe­rent levels of a Six-Sig­ma-pro­ject. 

How to App­ly Six Sig­ma? 
The strength of Six Sig­ma lies in the empi­ri­cal, data-dri­ven approach (and the use of quan­ti­ta­ti­ve mea­su­res of per­for­mance) to reach the goal of pro­cess impro­ve­ments and reduc­tions in varia­ti­on. In “Six-Sig­ma-Qua­li­ty-Impro­ve­ment-Pro­jects” the work is orga­ni­zed fol­lo­wing the Six-Sig­ma-DMAIC model: 

In the defi­ni­ti­on pha­se the goal and scope of a pro­ject are defi­ned, issues are coll­ec­ted that need to be tack­led to reach a hig­her (bet­ter) sig­ma level.

In this pha­se data is gathe­red about the cur­rent situa­ti­on to form a base­line of the pro­cess and to iden­ti­fy pro­blems.

Iden­ti­fi­ca­ti­on of the cau­ses of qua­li­ty issues and con­fir­ma­ti­on of the­se using data ana­ly­sis.

Imple­men­ta­ti­on of solu­ti­ons that have been deve­lo­ped based on the know­ledge gai­ned in the Ana­ly­sis-pha­se.

This pha­se ensu­res that the impro­ve­ments imple­men­ted in the pre­vious pha­se stay valid and acti­ve

Each of the­se steps makes use of spe­ci­fic ana­ly­ti­cal (quan­ti­ta­ti­ve) methods that are part of the over­all spec­trum of methods sug­gested for Six Sig­ma.  
Fur­ther infor­ma­ti­on regar­ding Six Sig­ma can be found in two books that dis­cuss the Six-Sig­ma-metho­do­lo­gy and its appli­ca­ti­on: Six Sig­ma: The Breakth­rough Manage­ment Stra­tegy (2000) by M. J. Har­ry and P. Schroe­der and The Six Sig­ma Hand­book (2001) by T. Pyz­dek. 

Sta­tis­ti­ca sup­ports the data coll­ec­tion and ana­ly­sis on every level of a Six-Sig­ma-pro­ject and can hence ser­ve as the ana­ly­ti­cal basis of Six-Sig­ma-initia­ti­ves and imple­men­ta­ti­ons for com­pa­nies of any size. The soft­ware has the fol­lo­wing func­tion­a­li­ties: 

  1. Com­pre­hen­si­ve set of Six-Sig­ma-tools like the Six-Sig­ma-cal­cu­la­tor, Six-Sig­ma-reports with inte­gra­ted gra­phics and cau­se-and-effect-dia­grams (Ishi­ka­wa) 
  2. A Six-Sig­ma-menu that is orga­ni­zed accor­ding to the “Six Sig­ma DMAIC”-approach. Addi­tio­nal Six-Sig­ma-rela­ted tools can be embedded in the UI 
  3. Cal­cu­la­ti­on of the pro­cess capa­bi­li­ty respec­ting the time-depen­dent dis­tri­bu­ti­on models accor­ding to DIN 55319 / ISO 21747 
  4. High qua­li­ty and fle­xi­ble gra­phing capa­bi­li­ties. The­se capa­bi­li­ties are also acces­si­ble via pro­gramming (in Sta­tis­ti­ca Visu­al Basic) which offers a near­ly unli­mi­t­ed poten­ti­al to cus­to­mi­ze and build tail­or-made visua­liza­ti­ons. 

Advan­ced Methods 
Sta­tis­ti­ca sup­ports advan­ced methods like Machi­ne Lear­ning and Data Mining algo­rith­ms to be appli­ed as alter­na­ti­ve approa­ches in Six-Sig­ma-pro­jects. The­se can be the foun­da­ti­on of an inno­va­ti­on advan­ta­ge over to the com­pe­ti­ti­on. 
Addi­tio­nal­ly, the workspaces pro­vi­de a visu­al inter­face to design and auto­ma­te ana­ly­ses. Using workspaces, it is pos­si­ble to share the work bet­ween all invol­ved team mem­bers easi­ly. For exam­p­le, a Black-Belt might design ana­ly­ses as a workspace to be later used by Green-Belts for appli­ca­ti­on in the field. 

Fur­ther advan­ta­ges are: 

  • Imple­men­ta­ti­ons of Machi­ne Lear­ning and Data Mining methods for clas­si­fi­ca­ti­on and pre­dic­tion inclu­ding arti­fi­ci­al neu­ral net­works, decis­i­on trees, sup­port vec­tor machi­nes and others. 
  • Algo­rith­ms spe­ci­fi­cal­ly sui­ted to work with lar­ge data sets 
  • Extra­c­tion of pro­cess rela­ted infor­ma­ti­on from his­to­ri­cal data using Fea­ture Sel­ec­tion as alter­na­ti­ve to expen­si­ve expe­ri­men­ta­ti­on. 
  • Workspaces to auto­ma­te ana­ly­ses and data pre­pa­ra­ti­on (fea­ture engi­nee­ring, clea­ning, fil­te­ring, sha­ping etc.) 

With Sta­tis­ti­ca Ser­ver it is pos­si­ble to design com­pa­ny-wide appli­ca­ti­ons for qua­li­ty assu­rance and impro­ve­ments using Six Sig­ma. The cus­to­miza­ti­on capa­bi­li­ties allow to shape Sta­tis­ti­ca into a tool that feels and acts like it was spe­ci­fi­cal­ly desi­gned for your needs. The plat­form pro­vi­des: 

  • Real-time moni­to­ring and alar­ming for the pro­duc­tion site, ana­ly­ti­cal tools for the engi­neer and report­ing opti­ons for the manage­ment. 
  • Cen­tra­li­zed set­up and manage­ment of data­ba­se queries and ana­ly­sis tem­pla­tes for shared use of data sources and spe­ci­fic appli­ca­ti­ons. 
  • User-spe­ci­fic inter­faces for all invol­ved par­ties: from simp­le UIs for the worker to hig­her methods for the Green-Belts to high­ly com­plex envi­ron­ments for ana­ly­ses and data mining for mas­ter Black-Belts. 
  • The plat­form is not only a com­pre­hen­si­ve envi­ron­ment for com­plex Six-Sig­ma-ana­ly­ses and data mining, but it also ser­ves well as a trai­ning envi­ron­ment. 
  • Ser­ver-based moni­to­ring of pro­ces­ses and qua­li­ty impro­ve­ments for an auto­ma­ted con­trol pha­se with fle­xi­ble alarms. 
  • Sca­le-able, cus­to­mizable, and easy to inte­gra­te into exis­ting data­ba­se and ERP-sys­tems. 
  • Extra­ct, Trans­form and Load (ETL) data from hete­ro­ge­nous data sources. 
  • Sche­du­ling can be used to pro­du­ce reports with up-to-date data. 
  • Exe­cu­ti­on of resour­ce inten­si­ve pro­cess on the ser­ver. 
  • Acces­si­ble via web brow­ser and can be tail­o­red to spe­ci­fic user needs. 
  • The inte­gra­ted cus­to­miza­ti­on opti­ons allow for tail­or-made web-based infor­ma­ti­on por­tals to allow access to cur­rent qua­li­ty ana­ly­ses of the Six-Sig­ma-pro­jects. 
  • Simul­ta­neous moni­to­ring of thou­sands of qua­li­ty mea­su­res on one or more ser­vers and easy manage­ment from a cli­ent PC. 
  • Cen­tra­li­zed instal­la­ti­on and manage­ment of your Six-Sig­ma-soft­ware. 

Stat­Soft as Your Part­ner 
Stat­Soft is your relia­ble part­ner when it comes to the soft­ware Sta­tis­ti­ca. We assist you with the sel­ec­tion, con­fi­gu­ra­ti­on and instal­la­ti­on and will also pro­vi­de you with the nee­ded soft­ware- and metho­do­lo­gi­cal-know-how. In ana­ly­ti­cal pro­jects we can aid with con­sul­ting and exe­cu­ti­on. Tog­e­ther we can gene­ra­te more insights from your data and deli­ver a sus­tainable value. 

Latest News
Your contact

If you have any ques­ti­ons about our pro­ducts or need advice, plea­se do not hesi­ta­te to cont­act us direct­ly.

Tel.: +49 40 22 85 900-0

Sasha Shiran­gi (Head of Sales)