Task 1 3 A SMART Decision Making Framework

Task 1. 3 „A SMART Decision Making Framework for BIM“ Annual Review Meeting Presented by Marcus Schreyer Max Bögl Bauservice Some slides are skipped for a presentation of this method at TNO Science and Technology. Frank Verhofstad, Technical Coordinator In. Pro February 2009 Paris, 13 -14 November 2008

Agenda q Introduction to the SMART DM Framework for BIM q SMART Framework Components q KPIs - Transforming Values and Objectives into KPIs q Priorities - Prioritizing Values with the Saaty Method q Representation - Visualizing evaluation results q Expected Application of T. 1. 3 Results within Inpro q Conclusion Annual Review Meeting Paris, November 2008 q 2

Introduction – Task Objective Design Alternative 1 BIM & Analysis Data Design Alternative 2 BIM & Analysis Data Design Alternative 3 BIM & Analysis Data Objective: Development of a formal, model-based decision making framework to enable a target value driven design Annual Review Meeting Paris, November 2008 q 3

Visualization of Results - Direct comparison of KPIs Rating Matrix Indoor Comfort Energy Consumption Cost Effective Alternative 1 100% 50% 100% Alternative 2 100% 80% 90% Alternative 3 100% 90% 80% Annual Review Meeting Paris, November 2008 q 4

Introduction - Required Framework Components Design Alternative 1 BIM & Analysis Data Design Alternative 2 BIM & Analysis Data Design Alternative 3 BIM & Analysis Data SMART DM Framework Evaluation Method Priorization Method KPI Functions Performance Visualization Stakeholder Values Annual Review Meeting Overall Alternative Ranking Paris, November 2008 q 5

SMART DMF Components SMART = Simple Multiattribute Rating Technique KPIs = Key Performance Indicators Priority Vectors Annual Review Meeting • Simple: The method should transparent and easy to use and not require complex explanations • Multiattribute: Designing is a task that generally tries to fulfill several, often contradicting goals at once • Rating: The design alternatives of options have to be evaluated and assigned a rating on a common, predefined scale • KPIs formally define the critial decision attributes or at Inpro- the critical design values • Method by the mathematician T. D. Saaty to set priorities in complex decision problems using Eigenvectors Paris, November 2008 q 6

KPIs – Definitions for an implementation into SMART „Client wants an energy efficient design“ „Client suggests Energy Class A“ Target Value Rating Scale KPI Utility Function Acceptance for Deviation from Target BIM Data „The regulations define energy classes A-F“ „Designers recommend classes A-B as good results, C to be considered as still acceptable“ „An energy analysis software calculates the energy class for a given design alternative from its BIM“ KPI components have to be selected and specified with a target value and utility function for each project! Annual Review Meeting Paris, November 2008 q 7

KPIs - Utility Function Example for Low Energy Use Rating Scale KPI Energy Certificate Class Utility Function Rating [%] 100 „With an energy class B, the BIM gets a performance rating of 75% for the criteria „energy class“ BIM Data 75 50 Criteria Scale: Dimension depending on the chosen criteria 25 Rating Scale: Degree of satisfaction 100% Fulfillment … 0% Fulfillment dimensionless Acceptance for Deviation from Target 0 A B C D E F G Energy Class „Thefunction: calculated energy class of the BIM Utility is class Defines the. B“relation between measurable criteria and the assigned rating. Also, in how far deviations are accepted. Defined by Client and development experts Utility functions establish the relation between the BIM of a design performance and the alternative Client/development team satisfaction Annual Review Meeting Paris, November 2008 q 8

Priorization – Method Requirements: § § Transparent Methodological Works with up to 15 items Numeric Method for priority setting in complex problems? Client and Development Team The pairwise comparison algorhythm by T. D. Saaty fulfills the requirements! Annual Review Meeting Paris, November 2008 q 9

Priorization - Overview on pairwise comparison by Saaty Evaluation Matrix P 1 = Indoor Climate Class 1 Energy Certificate Class 1/5 Return on Investment 1/3 9 Point Rating Scale by T. D. Saaty Annual Review Meeting Energy Certificate Class Return on Investment is equally important as 31 is moderately more 5 is strongly more important as as 3 5 1 3 1/3 1 Equally important Moderately more (or less) important Strongly more (or less) important Very strongly more (or less) important Overwhelmingly more (or less) important Paris, November 2008 The Evaluation Matrix is quadratic and reciprocal! 1 3 5 7 9 (or (or 1/3) 1/5) 1/7) 1/9) q 10

Priorization – Calculating the Priority Vector Numeric calculation of the highest Eigenvalue with the poweriteration (German: „Potenzmethode“) 1 st iteration: P 1 ² Rank sums Annual Review Meeting Eigenvalues r 1 Normalization Eigenvector w 1 Paris, November 2008 q 11

Priorization – Calculating the Priority Vector Numeric calculation of the highest Eigenvalue with the poweriteration (German: „Potenzmethode“) 1 st iteration: P 1 ² Rank sums Eigenvalues r 1 Normalization Eigenvector w 1 2 nd iteration: P 2 ² Rank sums Eigenvalues r 2 Normalization Eigenvector w 2 Checking the deviation of the Eigenvector: W 2 – W 1 0 if deviation is ~ 0 If it‘s >> 0 Annual Review Meeting w 2 = Priority Vector Continue iteration Paris, November 2008 q 12

Visualization of Results - Direct comparison of KPIs Rating Matrix Indoor Comfort Energy Consumption Cost Effective Alternative 1 100% 50% 100% Alternative 2 100% 80% 90% Alternative 3 100% 90% 80% Ratings taken from the utility functions Annual Review Meeting Paris, November 2008 q 13

Visualization of Results - Total weighted ratings Indoor Climate Class Energy Certif. Class Alternative 1 100% 50% 100% Alternative 2 100% 80% 90% Alternative 3 100% 90% 80% Rating Matrix Annual Review Meeting Cost Effective Priority Vector Total Rating 94, 8 % 63, 70 X 10, 47 = 25, 83 Paris, November 2008 95, 3 % 93, 8 % q 14

Conclusions q A framework for making value driven design decisions has been developed q Another application could be the selection of bidders during procurement in a multicriteria evaluation process (not just by the lowest bid) q The SMART DMF has already successfully been demonstrated in a „proof of concept“ study in a collaboration with T. 2. 3 and 2. 4 see Application Scenario later on by Prof. Thomas Olofsson q The SMART DMF method has already been coded as a MS Excel-tool by Lulea University to be used for the key processes. Therefore the method can fairly easily be implemented into a software solution. All Task Objectives successfully fulfilled! Annual Review Meeting Paris, November 2008 q 15

Participants of Task 1. 3 q Max Bögl Bauservice Gmb. H (Task Leader) q P 3 – Digital Services Gmb. H q University of Dortmund q Bouygues Traveaux Publics SA q HOCHTIEF AG q Swedish Association of Building Construction Clients Annual Review Meeting Paris, November 2008 q 16

q. This document may not be copied, reproduced, or modified in whole or in part for any purpose without written permission from the In. Pro Consortium. In addition to such written permission to copy, reproduce, or modify this document in whole or part, an acknowledgement of the authors of the document and all applicable portions of the copyright notice must be clearly referenced. q. All rights reserved. This document may change without notice. q. Disclaimer: The information in this document is provided as is and no guarantee or warranty is given that the information is fit for any particular purpose. The user thereof uses the information at its sole risk and liability. Annual Review Meeting Paris, November 2008 q 17