Model-based decision processes for agenda building and project funding Eeva Vilkkumaa Lectio praecursoria The document can be stored and made available to the public on the open internet pages of Aalto University. All other rights are reserved.
Intelligent products Business concept development Future resources Improvement of energy efficiency Big data Self-organizing systems Print-on-wood Aesthetic features Efficient substitution Integrated Billing E-energytrading.com CyberRetail.com E-notification Outsourced energy management Online environmental management E-biz in a box
One million No. of people on earth No. of insects on earth
Decision model Decision recommendations Multiple stakeholders Uncertainties Risk attitudes Objectives Values
YES! NO! A B C D E F G H I J K L M Value increases for stakeholder 1 Value increases for stakeholder 2
60%!! A B C D E F G H I J K L M 40%... Precise importance weights → Single recommended portfolio
>25% A B C D E F G H I J K L M = Core project – agree to select = Exterior project – agree to reject = Borderline project – negotiate Incomplete importance weights → Set of non-dominated portfolios
1. Digital economy of global networks Technology and markets rule Huge global growth 2. Adaptation to scarcity Strong regulation Renewable resources, recycling and resource effiency 4. World of conflicts Superpowers and realpolitik rule Tension between cultures 3. Global economic crisis Tax evasion, corruption and black economy Infrastructure and public services deteriorate quickly Global scenarios for 2025 Source: Huoltovarmuuden skenaariot 2025, Huoltovarmuuskeskus, A: International cooperation against tax havens B: Increased armament F: Investment in renewable energy technologies % 35% 25% 10% ʻ Scenario 2 is more probable than scenario 1’ 28% 37% 23% 12% A: International cooperation against tax havens? B: Increased armament? C: Decreased armament? D: Protectionism? G: Free trade agreements? H: Leave the EU? I: Abolish euro? F: Investment in renewable energy technologies? E: Investment in nuclear energy? ʻ Selection of project B makes the probability of scenario 4 higher than 15%’
A B C D E F G H I J K L M = Core project – select regardless of what happens = Exrterior project – reject regardless of what happens = Borderline project – wait and see Incomplete scenario probability information → Set of non-dominated portfolios
The inevitable Dilbert strip
E D CB A F True value of each project Distribution of estimation error True cost of each project A B EF CD FE AB
Source: Flyvbjerg et al. (2002), Underestimating Costs in Public Work Projects – Error or Lie? Journal of the American Planning Association, Vol. 68, pp Average escalation μ=27,6% Cost escalation (%) Frequency (%)
AA Value estimate for candidate project Adjusted estimate for candidate project Prior distribution of the values of similar projects Prior mean Distribution of estimation error
D D A B C E F G H I J Estimated value (M€) Projects E Selection threshold E Select Re-evaluate and reconsider Reject
Value estimate for candidate project Adjusted estimate for candidate project Prior distribution of the values of similar projects Prior mean Distribution of estimation error A
Photocopier Personal computer Gene targeting of mammalian cells Idea rejected in 1970s by 17 companies such as Xerox, IBM, and Hewlett Packard – the rest is history Received unanimously negative reviews at National Institutes of Health in 1980s – Nobel Prize in 2007 Rejected by Kodak as a bad idea in 1940s – later made Xerox a huge success
To fund exceptionally valuable projects To maximize short-term performance FF = full funding R = rejected projects CF= conditional funding R = rejected projects C = continued projects A = abandoned projects