Portfolio decision analysis for infrastructure and innovation management

Pekka Mild

Research output: ThesisDoctoral ThesisCollection of Articles

Abstract

Practically all companies and public agencies make decisions about resource allocation among investment categories and selection of 'lumpy' investments such as projects. While economic calculations tend to be central in these decisions, it is often necessary to account for hard-to-monetize impacts, multiple objectives, stakeholder interests and relevant constraints. For these decisions, Portfolio Decision Analysis (PDA) provides mathematical methods and decision support processes that assist in the construction of portfolios consisting of a set of discrete alternatives. This Dissertation develops a multi-criteria PDA methodology, Robust Portfolio Modeling (RPM), for problems in which a subset of indivisible projects is to be selected from a large number of candidates, with the aim of contributing to the attainment of multiple objectives while satisfying relevant budgetary and other constraints. The RPM methodology – which is based on the linear-additive value representation – admits incomplete information about criterion weights, project scores and project costs in portfolio selection. A key concept of the RPM, the core index, derives project-level recommendations by computing and analyzing non-dominated project portfolios. These recommendations show (i) which projects are robust choices in the light of the incomplete information and (ii) which projects are promising targets for acquiring additional information that can lead to more conclusive results concerning their inclusion or exclusion to the final portfolio. The Dissertation also presents real-life applications of PDA in infrastructure and innovation management. The RPM application on bridge maintenance management (Paper [III]) found its way into repeated use at the Finnish Transport Agency. The second infrastructure application (Paper [IV]) is an innovative combination of standard Operations Research methods to support strategic resource allocation between road asset categories and types of operations. The third application (Paper [V]) demonstrates how the RPM can be utilized to ex post evaluation to identify sets of over- and underperforming projects in an innovation program. The Dissertation shows that the RPM method and its key concepts are readily understood and accepted by practitioners, including senior managers. The method can be tailored to utilize existing project data from sources such as monitoring databases, and it is computationally capable of processing hundreds of project candidates. Moreover, given that repeated real-life applications are relatively rare in decision analysis literature, the methodological development in this Dissertation already can be viewed as a pioneering platform for further avenues of PDA research.
Translated title of the contributionPortfoliopäätösanalyysi infrastruktuuri- ja innovaatiojohtamisessa
Original languageEnglish
QualificationDoctor's degree
Awarding Institution
  • Aalto University
Supervisors/Advisors
  • Salo, Ahti, Supervising Professor
  • Salo, Ahti, Thesis Advisor
  • Liesiö, Juuso, Supervising Professor
Publisher
Print ISBNs978-952-60-7282-1
Electronic ISBNs978-952-60-7281-4
Publication statusPublished - 2017
MoE publication typeG5 Doctoral dissertation (article)

Keywords

  • portfolio decision analysis
  • project prioritization
  • resource allocation
  • multi-attribute value theory
  • incomplete information
  • robustness
  • applications
  • infrastructure asset management
  • innovation management

Fingerprint

Dive into the research topics of 'Portfolio decision analysis for infrastructure and innovation management'. Together they form a unique fingerprint.

Cite this