Bismark Singh

Interdisciplinary Research

My primary academic interest lies in advancing data-driven decision-making under uncertainty, particularly through two-stage stochastic optimization. This involves both theoretical innovation and practical application, addressing pressing societal and environmental challenges by formulating them as data-driven mathematical optimization models. These models are inherently large-scale, requiring the development of modern algorithms to enable computationally efficient solutions.

A central focus of my research is on chance-constrained optimization, where the goal is to design systems that remain resilient even under extreme risk conditions. I work on developing specialized algorithms to solve such stochastic programs, ensuring tractability and practicality.

Motivations and Applications

My work is inherently interdisciplinary, driven by real-world challenges such as:

Energy Systems: Designing resilient and sustainable energy infrastructures under uncertain demand and supply.
Pandemic Response: Guiding healthcare policy to ensure fairness in populations and allocate scarce resources effectively.
Critical Risk Management: Minimizing socioeconomic impact of unforeseen man-made attacks or natural disasters.
Sustainable Waste Management: Developing models for efficient positioning of recycling centers for environmental sustainability.

A unique aspect of my approach is modeling subjective human behavior within the decision-making process. This adds a layer of complexity and realism to my models, bridging the gap between mathematical rigor and human-centric decision-making.

Probabilistic Bounds

This research represents a theoretical interest where I explore joint-chance constraints through the lens of classical probability theory. Specifically, I approach a joint-chance constraint as a union of sets and utilize classical probability bounds to constrain it effectively. Bounding the probability of the union of n events using joint probabilities of k < n events has a rich history, dating back to the foundational work of Boole and Bonferroni. What makes this line of probabilistic research particularly interesting is its application in optimization models. When these probabilistic bounds are incorporated into chance-constrained optimization models, they provide upper and lower bounds on the optimal objective function value, providing both computational and theoretical insights. [Click for more →]

This interest originated during my PhD studies and matured further following my first major grant as a Principal Investigator during my position at Sandia National Labs, US (2018). The grant supported significant advancements in this domain, culminating in two publications: here and a follow-up here. Currently, this work is being extended collaboratively with my PhD student, focusing on deeper theoretical insights and broader applications.

Satisfying a joint chance constraint is an intersection of "successes". Explore→

Several European countries are investing on renewable energy sources.

Sustainability & Climate Change

As countries advance towards Net-Zero goals (e.g., Germany's Energiewende), joint chance constraints have proven especially effective for ensuring the highly reliable operation of critical energy systems under uncertain renewable energy availability. Key examples include photovoltaic (PV) systems and coupled wind-diesel systems. Mathematical optimization models for such systems often present significant challenges due to their structure and scale, necessitating the development of novel algorithmic approaches.
In my research, we have designed modern machine-learning-inspired iterative algorithms and employed Lagrangian-based heuristics to tackle these challenges. See examples here and here. [Click for more →]

At Sandia National Labs, US (2016–19), I focused on solving large-scale energy system models, addressing critical risks faced by the US electrical grid. Many of these works are available on the US Department of Energy's Office of Scientific and Technical Information's website. Access here →
At FAU Erlangen-Nürnberg, Germany (2019–22), I led the chair’s contributions to the multi-institute “METIS” research collaboration with the Jülich Research Center. This project develops open-source tools for optimizing large-scale energy system models under the framework of Germany’s Energiewende (transition to clean, yet reliable energy systems).
- Learn more about the project here →
- Explore the technical details here →

This research not only advances mathematical optimization but also contributes to global sustainability goals, ensuring renewable energy systems remain both efficient and reliable under uncertainty.

Fairness & Waste Management

In 2020, I began exploring undesirable facility location problems (FLPs) with a unique perspective: ensuring fairness from the point of view of the facilities, rather than the users, which is the typical approach in existing literature. These models have significant applications in waste management, particularly in the context of recycling centers (e.g., UK “tips” or German Wertstoffhöfe). Recycling centers, which allow populations to dispose of certain recyclable waste, are critical for achieving sustainability goals. However, they also contribute to pollution and public nuisance, leading many governments to consider their closure. How can such decisions be made ethically and fairly?

Supported by the Bavarian State Ministry for Science & Arts and the University of Southampton, my team developed discrete optimization models that achieve fairness in facility closures. Key milestones include:

Supervising two student theses on this topic, resulting in published articles.
Leading a postdoctoral researcher in quantifying subjective opinions on recycling campaigns, incorporating human perceptions into decision models. [Click for more →]

Theoretical: Defined new axioms of fairness, shifting the perspective to the facilities themselves. Formulated novel classes of FLPs with interesting mathematical properties, particularly in their Karush-Kuhn-Tucker (KKT) optimality conditions. Explore→
Computational: Addressed the intractability of solving these models naively by designing specialized algorithms, enabling efficient solutions for large-scale instances of the size of Bavaria and all of Germany. Explore→
Societal: Ensured sustainability goals are achieved without disproportionately affecting certain communities by offering governments ethically fair decision-making tools for closing recycling centers while maintaining public accessibility. Explore→

The Bavarian government is shutting down recycling centers in the past two decades. Explore→

The system of staged lockdowns used by Austin, Texas during the COVID-19 pandemic designed by our collaborative work. Explore→

Pandemic Risk Mitigation

I began collaborating with the Texas Department of State Health Services, US, as a MSc student in 2012, helping prepare for future pandemics long before the emergence of COVID-19. Motivated by Texas' response to the 2009 H1N1 pandemic, my PhD research focused on designing web-based, optimization-backed decision-support tools for government use. These tools, accessible at flu.tacc.utexas.edu, assist the State of Texas in the fair and efficient allocation of critical resources, such as antivirals and vaccines. [Click for more →]

During the COVID-19 pandemic, I renewed this collaboration to address pressing challenges in pandemic response. Key contributions include:

Testing Accessibility: Measuring the reach and equity of COVID-19 testing across the United States. Explore→
Triggering Lockdowns: Designing the staged-dashboard lockdown system employed by the City of Austin, which integrates a chance-constrained framework for informed, adaptive public health policies. Explore→

Research Funding and Impact

Most significant funding is provided by a 3-year grant from the Deutsche Forschungsgemeinschaft (German Research Foundation), with additional funding from the Bavarian Czech Academic Agency and the Horizon 2020 Program. See Publications.

Google Sites

Report abuse