Public defence, Mathematics and Statistics, MSc Joaquín de la Barra Toloza

Title of the thesis: Models of Portfolio Decision Analysis for Mitigating Risks in Critical Networks

Thesis defender: Joaquín de la Barra Toloza 
Opponent: Professor Enrico Zio, Politecnico Milano, Italy
Custos: Professor Ahti Salo, Aalto University School of Science

Modern societies rely on critical infrastructure such as electricity grids, railway networks, water distribution systems, and telecommunications networks to provide essential services. Natural hazards, technical failures, human error, and malicious attacks can damage this infrastructure and disrupt its services. Infrastructure managers must therefore decide how to use limited resources to prevent failures, strengthen vulnerable components, and restore services efficiently after disruptions.

This dissertation develops methods to help infrastructure managers make these decisions. For example, the methods can help determine which parts of an electricity grid or railway to strengthen and the order in which to restore damaged railway components. They account for the fact that future disruptions cannot be predicted with certainty and that improving one aspect of a system may require compromises in another. The research extends existing approaches by combining methods for assessing infrastructure performance with methods for comparing alternative actions and allocating limited resources.

For electricity distribution grids, the dissertation develops a framework to identify cost-efficient combinations of actions to strengthen the grid. The framework supports decisions even when the likelihood of different hazards or the relative importance of different reliability objectives is only partially known. This allows comparing alternative investments without requiring precise information that may be difficult to obtain in practice.

For railway systems, the dissertation develops methods to identify the most critical components for maintaining services and determine where investments can most effectively improve service reliability and resilience. It also examines how to prioritize disrupted components for restoration to minimize overall service loss during a disruption. The methods allow consideration of multiple services and objectives when allocating limited resources.

The dissertation also examines how authorities can disrupt the supply of counterfeit microchips. Counterfeit chips can enter legitimate supply chains and potentially affect critical infrastructure that depends on electronic components. Because resources for combating counterfeit trade are limited, authorities must decide which measures to use and where to focus their efforts. Suppliers and buyers may also adapt their behavior in response to these measures. The dissertation develops a model that accounts for these reactions and helps identify combinations of measures that can reduce counterfeit microchip trade.

Overall, the dissertation shows how risk assessment, decision analysis, and mathematical optimization can be combined to support decisions about strengthening and restoring critical infrastructure and mitigating threats to its operation. A key contribution is that the developed methods can support decisions even when information about future hazards or the relative importance of different objectives is incomplete. Infrastructure managers, system operators, and policymakers can apply the methods to compare alternative actions and direct limited resources toward measures that improve the reliability and resilience of essential services.

Key words: critical infrastructures, probabilistic risk assessment, portfolio decision analysis, equilibrium models

Thesis available for public display 7 days prior to the defence at Aalto University's public display page. 

Contact: joaquin.delabarra@aalto.fi