TEAM FORMATION. Outline. QFD in Team Formation. Quality Function Deployment (QFD) Basic Planning Matrices

Transcription

1 TEAM FORMATION Adrew Kusiak Seamas Ceter Iowa City, Iowa - Tel: - Fax: - adrew-kusiak@uiowa.edu Outlie Itroductio The quality fuctio deploymet methodology The aalytical hierarchy process approach The mathematical programmig model Model extesio Summary The cocept of multi-fuctioal teams is importat aspect of performig busiess activities ad problem solvig i may idustries The cocept of multi-fuctioal teams is oe of the key aspects i product developmet - Experts from various disciplies, such as, desig, maufacturig, quality testig, ad marketig work i groups rather tha idividually i order to develop a quality product - The membership of a team depeds o the type of product to be developed, customer requiremets, ig ad product, ad so o Quality Fuctio Deploymet (QFD) The essece of QFD is to trasform customer requiremets ito "quality " ad develop product desigs of high quality by systematically deployig the relatioships betwee customer requiremets ad ig. The QFD emphasizes: ) customer ) customer requiremets ) how to meet the customer requiremets Basic Plaig Matrices (a) (b) QFD i Team Formatio Fuctioal team members Customer requiremets A vehicle for: represetatio of depedecies data collectio (a) customer requiremets - ig plaig matrix (b) ig - team members plaig matrix

2 EXAMPLE Hierarchical structure for team formatio Level Customer requiremets Level Level Fuctioal teams Goal k m Remember A egg characteristic represets A set of activities These activities may form a project A multifuctioal team is formed to perform these activities Mathematical Programmig Model Notatio i - ig characteristic idex j - fuctioal team member - umber of ig m - umber of fuctioal team types wij - weight correspodig to fuctioal team j resposible for realizig ig characteristic i p - umber of multi-fuctioal teams mj - umber of that ca be hadled by a team member of fuctioal team type j. The value of mj is a fuctio of time, techological ad schedulig costraits, etc. M - a arbitrary large positive umber xij = if a team member of fuctioal team type j belogs to the multidiscipliary team realizig ig characteristic i = 0 otherwise yi = if a multidiscipliary team for ig characteristic i is formed = 0 otherwise activities The Model? Formig Multi-fuctioal Teams m Max wijx ij i= j= s. t. xij m j j =,..., m i= yi p Max No. of iterdiscipliary teams i= m xij My i i =,..., j= i - ig characteristic idex (or set of desig activities) j - the type of a fuctioal team Max No. of (activities) to be hadled by a member of fuctioal team j Cosistecy costrait (yi = if team for characteristic i is formed) x ij = 0, i =,..., j =,..., m y i = 0, i =,..., Two ways of geeratig weights wij: Direct geeratio of mx matrix characteristic (or Set of activities).. Fuctioal team type 0 Aalytical Hierarchy Process

3 Case Study Aalytical Hierarchy Process (AHP) Methodology The multicriteria decisio makig leads to the selectio of the best alterative uder coflictig criteria AHP process - Costruct a hierarchy - Determie the priorities (importace measures) of the elemets at each level of hierarchy - Sythesize priorities to determie the priorities of decisio alteratives - Costruct a compariso matrix - Perform pairwise compariso of each pair of customer requiremets with respect to the overall goal Scale of Relative Importace Modified for Team Selectio,,, 8 Defiitio Equal importace Moderate importace oe over aother Essetial or strog importace Very strog or demostrated importace Extreme importace Itermediate values betwee two adjacet judgmets Reciprocal Explaatio Two views cotribute equally to the goal Experiece ad judgmet moderately favors oe view over aother Strogly favors oe view over aother A view is very strogly favored ad its domiace is demostrated i practice The evidece of favorig oe view over aother oe is of the highest possible level of affirmatio Whe a compromise is eeded For iverse comparisos EXAMPLE: Automotive product Customer requiremets - ig plaig matrix Customer Requiremets Characteristics Eve brakig Easy brake fluid replacemet Normal fuctioig of the brakes after egie stops ruig Easy brake replacemet Lighter brake pedal pressure Automatic trasmissio No erratic shiftig 8 No delay egagemet No trasmissio overheatig 0 Easy oil replacemet Fuel efficiecy Easy oil check ad replacemet Quiet egie operatio Excellet egie performace Passeger side airbag Roomy back seats Roomy truk Doors easy to ope, easy to 8 close, easy to lock Corrosio protectio 0 Bright drivig beam No lightig troubles Brakes hydraulic system Self adjustig brakes Torque coverter Plaetary gears ad cotrols Egie lubricatig system Cylider block ad head Combustio chamber Valves ad portig Safety Coveiece Circuit breakers - team members plaig matrix Characteristics Fuctioal Team Members Mechaical Maufacturig Quality Fiace expert Electrical Reliability Hierarchy for the Model i the Example Goal Build a car Brakes hydraulic system Self adjustig brakes Torque coverter Plaetary gears ad cotrols Egie lubricatig system Cylider block ad head Combustio chamber Valves ad portig Safety Coveiece Circuit breakers Level Customer requiremets Level Level Team members Mechaical Maufactur. Quality Fiace expert Electrical Reliability

4 Pairwise Compariso of Customer Requiremets With Respect to the Goal / / / / / / / / / /aij / / / / / / / / / / / / / / / /8 / / /8 / / / / / / λmax =.080 CI =0.0 CR =0.08 / / / / / / / / / / / / / / / / / / /8 / / / / / / / / / / / / / / / /8 / / / / / / / / / / / / / / / / / / / / / / / / / / / / / / / / / / / / / / / / / / / / / / vector (W) Eigevalues ad Eigevectors EXAMPLE A = I = To obtai eigevalues A-λI = -λ -λ 0 0 λi = = (- λ)( - λ) - = (λ - λ ) = 0 λ 0 0 λ The eigevalues W = λ, λ = + λ = To obtai eigevector (priority vector) W correspodig to λ AW =λw w w = w λ w w + w =λw w + w =λw W =, (here for λ max λ) We ca ormalize W by equatig its coefficiets total to uity. It is doe by dividig each coefficiet by the sum w +w which is λ + λ The resultat ormalized vector is, λ + λ λ + The cosistecy idex (CI) ad cosistecy ratio (CR) for the compariso matrix A = [aij] are computed as follows: CI = ( λ max - )/( - ) CR = (CI/ACI) 00% where:- λ max is the largest eigevalue of the compariso matrix, - is the dimesio of the matrix, - ACI is the average cosistecy idex for radomly geerated weights (from Saaty s table) The relative preferece throughout the hierarchy is obtaied performig pairwise comparisos, where a decisio maker or group of decisio makers express their judgmets For cosistet comparisos: The cosistecy ratio CR < 0%

5 Eigevector A vector represetig a matrix Determiig wi s Eigevalue way A W From the table Eigevalue way The priority vector of customer requiremets ca determied usig the largest eigevalue λmax of matrix A. The eigevector correspodig to the λmax is the priority vector of customer requiremets. The computatio for the data i matrix A yields the priority vector W with the largest eigevalue λmax =.080, the CR is computed. The cosistecy idex is CI = 0.0 ad the cosistecy ratio CR is Note, the value of average cosistecy idex ACI =. is used which was obtaied for radomly geerated weights (Saaty 8). From the table Priorities of Customer Requiremets The weight wi of row i of matrix A = [aij] is computed as follows: wi = Total of aij i row i / Total of aij for all i ad j Customer requiremet

6 Goal Build a car Weight Geeratig Priciple Level Customer requiremets Attribute; weight wi Level Top dow Multiply values by wi Aggregate to get wj Level Team members Mechaical Maufactur. Quality We were at this level Fiace expert Electrical Reliability Attribute; weight wj Based o the followig tables tables x tables x tables x tables x tables x Table of Normalized Weights of Characteristics Brakes hy draul i c sys tem Self adjustig brakes Torque coverter Plaetary gears ad cotrol Egie lubricatig system Cylider block ad head Combustio chamber Valves ad portig Safety Coveiece Circuit brakers Weight vector (wi) Total = x (Collect data) Req. -Eve breakig x Multiplyig values by wi w x x x x x w x x w w x Req. - Light brake pedal pressure w Matrix aggregatio Add all x vectors to geerate oe x vector x The result - weights Brakes hy draul i c sys tem Self adjustig brakes Torque coverter Plaetary gears ad cotrol Egie lubricatig system Cylider block ad head Combustio chamber Valves ad portig Safety Coveiece Circuit brakers Vector (wi)

7 Bar Chart of Normalized Weights of Characteristics Level Customer requiremets Goal Build a car Level Level Team members Mechaical Maufactur. Quality Fiace expert Electrical Reliability characteristic We were at this level To be doe ext Example: Oe of 8 matrices Egg characteristic - Brakes hydraulic system Based o 8 x tables, the matrix o the ext slide is geerated. fuctioal areas of team members w=0.08 x (Collect data) Wr The first row of the fial matrix Table of Normalized Weights of Team Members with Respect to Characteristics ME MF DE QE FE EE RE W W W W W W W Brakes hydraulic system Self adjustig brakes Torque coverter Plaetary gears ad cotrols Egie lubricatig system Cylider block ad head Combustio chamber Valves ad portig Safety Coveiece Circuit breakers Bar Charts of Normalized Weights of Team Members with Respect to Characteristics Brakes hydraulic system Self adjustig brakes Torque coverter Plaetary gears ad cotrols characteristic Egie lubricatig system Cylider block ad head Combustio chamber Valves ad portig characteristic ME MF DE QE FE EE RE ME MF DE QE FE EE RE

8 Bar Chart of Normalized Weights of Team Members with Respect to Characteristics (cotiued) EXAMPLE Data i additio to the weight matrix ME MF DE QE FE EE RE Fuctioal team type ME MF DE QE FE EE R E Number of Safety Coveiece Circuit breakers characteristic Max umber of teams = 8 max m wijx i= j= ij The Model i - ig characteristic idex j - type of a fuctioal team s. t. xij m j j =,..., m Max No. of i= to be hadled by a member of fu. team j yi p Max No. of iterdiscipliary i= teams m Cosistecy xij My i i =,..., costrait (yi = if j= team i is formed) x ij = 0, i =,..., j=,..., m y i = 0, i=,..., The Solutio x = x = x =, x = x = x = x = 0 x = x =, x = x = x = x = x = 0 x = x = x = x =, x = x = x = 0 x = x = x = x = x =, x = x = 0 x = x = x = x = x = x =, x = 0 x = x = x = x = x = x =, x = 0 x = x = x = x =, x = x = x = 0 x8 = x8 = x8 = x8 = x8 =, x8 = x8 = 0 x = x = x = x = x = x = =, x = 0 x0 = x0 = x0 = x0 = x0 = x0 =, x0 = 0 x = x = x = x = x = x =, x = 0 x = x = x = x = x = x = x = x = x = x = x = x = x = x = x = x = x = x = x = x = x = x = x = x = x = x = x = x = x = x = x = x = x = x = x = x = x = x =, x = x = x = x = 0 x8 = x8 = x8 = x8 = x8 =, x8 = x8 = 0 i= multidiscipliary team umber j= fuctioal team umber Teams Selected T = {ME, DE, RE} T = {ME, RE} T = {ME, MF, DE, RE} T = {ME, MF, DE, EE, RE} T = {ME, MF, DE, QE, EE, RE} T = {ME, MF, DE, QE, EE, RE} T = {ME, MF, DE, RE} T8 = {ME, MF, DE, EE, RE} T = {ME, MF, DE, QE, EE, RE} T0 = {ME, MF, DE, QE, EE, RE} T = {ME, MF, DE, QE, EE, RE} T = {ME, MF, DE, QE, FE, EE, RE} T = {ME, MF, DE, QE, FE, EE, RE} T = {ME, MF, DE, QE, FE, EE, RE} T = {ME, MF, DE, QE, FE, EE, RE} T = {ME, MF, DE, QE, FE, EE, RE} T = {DE, EE, RE} T8 = {ME, DE, FE, EE, RE} Modified Hierarchy for the Model i Example (Reduced umber of iteractios) Level Customer requiremets Level Level Compoets Level Team members Brakes Trasmissio Egie Mechaical Maufactur. Goal Build a car Quality Fiace expert Electrical Reliability Lightig system

9 Modified Example To reduce the umber of the AHP compariso matrices ad the size of the model, oe ca modify the hierarchical structure at level (Level ) Data for the Modified Example Normalized Importace Measure of Team Members with Respect to Subsystems ME MF DE QE FE EE RE w w w w w w w Power brakes Trasmissio Egie Lightig system Fuctioal team type ME MF DE QE FE EE RE Number of The Solutio x =, x =, x =, x =, x = 0, x =, x = x =, x = 0, x =, x = 0, x = 0, x =, x = x =, x =, x =, x =, x =, x =, x = x =, x =, x =, x =, x =, x =, x = x = 0, x = 0, x =, x = 0, x =, x =, x = 0 Teams T = {ME, MF, DE, QE, EE, RE} T = {ME, DE, EE, RE} T = {ME, MF, DE, QE, FE, EE, RE} T = {ME, MF, DE, QE, FE, EE, RE} T = {DE, FE, EE}