The Research Facility Support Grants (RFSG) competition provides funding for the operation, maintenance, replacement, repair or upgrade of equipment used in UBC facilities to support research. It intended to enhance UBC research, training, and equipment sustainability. RFSGs provide up to $50,000 for a term of one year (1) to provide one-time bridge funding to support staff in shared facilities or (2) to improve the capabilities of shared equipment. Researchers located on the UBC Vancouver campus and affiliated hospital sites are eligible to apply.
Social-Behavioral Drivers of Household Resource Consumption
Renewable Energy Transition
Equitable Climate Change Mitigation & Adaptation
Current Research Work
My research explores social dimensions of climate change in the US and Canada. I critically evaluate top-down and bottom-up approaches that aim to advance equitable climate change mitigation and adaptation efforts. With a focus on local planning and politics, I ask questions about decision making, governance, and collective action. My scholarship draws from interdisciplinary social science theories and methods and I prioritize working collaboratively with people impacted by climate change and the plans and policies developed to mitigate its impacts.
Predictive and Constrained Control: All systems are constrained, e.g., aircraft are subject to angle-of-attack constraints to prevent stall, autonomous vehicles must avoid obstacles, and electric motors have torque and power limits. I’m interested in developing control algorithms that account for these constraints in a systematic and optimal manner. In particular, I work extensively on model predictive control and reference governors.
Real-time Optimization: Solving optimization problems in real-time and in resource constrained environments is a critical for enabling complex goal-oriented behaviors in autonomous systems. I’m interested in developing optimization algorithms that are suited for deployment on embedded computers as well as distributed algorithms for multi-agent systems with limited communication capabilities.
Game-theoretic Control: Many critical engineering systems such as energy grids, traffic networks, or supply chains are made up of multiple interacting subsystems controlled by various human or automated agents. These agents rarely agree on system-wide objectives but still need to coordinate to manage shared resources and infrastructure (e.g., power generation, production capacity, or transit links). I am interested in developing coordination and control mechanisms that enable the safe coexistence of agents with conflicting objectives using tools from control and game theory.
Applications in Energy, Manufacturing, and Robotics: My research is applicable to a wide variety of application domains. I’m currently interested in applying optimization-based and game-theoretic control to additive manufacturing (3D-printing) processes, smart grids and buildings, precision motion systems, supply chains/logistic networks, and multi-agent planning for autonomous vehicles. In the past, I’ve worked on engine emissions control, landing spacecraft on asteroids, and upset recovery in aircraft.
Data Valuation and Auditing for GenAI: Data valuation plays a crucial role in machine learning. Existing data valuation methods, mainly focused on discriminative models, overlook generative models that have gained attention recently. In generative models, data valuation measures the impact of training data on generated datasets. We formulate the data valuation problem in generative models from a similarity matching perspective to bridge the gaps. Specifically, we introduce GMValuator, the first training-free and model-agnostic approach to providing data valuation for generation tasks.
Collaborative Learning via Textual Gradient on LLM: This project introduces Federated Textual Gradient (FedTextGrad), a novel approach that integrates textual gradient optimization into federated learning (FL). Instead of numerical gradients, FedTextGrad enables FL clients to refine prompts using LLM-generated textual feedback, with aggregation via text summarization. The study highlights key challenges in information retention and proposes an improved summarization method using the Uniform Information Density (UID) principle to enhance prompt effectiveness. Experiments on reasoning tasks validate the feasibility of this approach, expanding FL’s applicability to optimizing LLMs in decentralized settings.
FairMedFM: Fairness Benchmarking for Medical Imaging Foundation Models: FairMedFM, a benchmarking framework for evaluating fairness in medical imaging foundation models (FMs). It integrates 17 datasets, 20 FMs, and multiple fairness metrics to analyze biases across classification and segmentation tasks. The study highlights bias prevalence, dataset-specific disparities, and ineffectiveness of existing mitigation strategies. FairMedFM provides an open-source and extensible pipeline, aiming to enhance fairness evaluation and promote equitable AI applications in medical imaging.
Multimodal Medical Data Analysis with Graph Prompt Learning: This project introduces MMGPL, a graph prompt learning framework for diagnosing neurological disorders using multimodal medical data. MMGPL reduces irrelevant patches and leverages graph-based prompts to capture structural relationships among brain regions. By integrating GPT-4-generated disease concepts with a Graph Convolutional Network (GCN), MMGPL enhances medical data analysis. Experiments on neurological disorder datasets demonstrate its superior performance and scalability.
Multi-Agent Collaborative Decision Support System for Healthcare (MAC-Health): Performing diagnosis and helping patients manage treatment require doctors to use complex reasoning and decision-making, which often requires experts to make the decisions collaboratively. In recent years, large language models (LLMs) have achieved significant advancement, prompting the exploration of developing decision support systems for doctors to help improve clinical care. This has resulted in the emergence of a plethora of multi-agent systems (MAS) with foundation model-based agents providing interactive decision support based on queries about patient conditions. In this project, we develop MAC-Health as an AI copilot for doctors or medical school students.
Eni, in the framework of the 2025 Edition of the Eni Award Prizes, which aim to encourage a better use of conventional energy sources, to promote the gradual decarbonization of the energy system through the use of renewable sources, as well as to encourage research on environmental protection and to help new generations of researchers to emerge, announces:
THREE MAIN PRIZES * Advanced Environmental Solutions * Energy Transition * Energy Frontiers
Each Prize will consist of a customized plaque and an indivisible sum of €200,000 (Two hundred thousand Euros) and the participation in the prize giving Ceremony in Italy. Eni will grant hospitality for winners participating in the Ceremony plus a maximum of one accompanying person.
Please, note that the candidature procedures are totally free of charge and easily managed through a web platform. Potential candidates who wish to receive information or to apply may contact Laura Sterli (laura.sterli@feem.it).
Tyler Lewis Clean Energy Research Foundation has an opportunity for students who are Canadian or permanent residents. The “Tyler Lewis Clean Energy Research Grant” is awarded annually, at a value of $10,000. Depending on the caliber and worthiness of applicants and projects, the Foundation will award as many as three $10,000 grants each year.
The broad area of interest (clean energy research) allows for a diversity of applications– our only barrier to having a healthy competition is inspiring potential applicants to apply. If your research fits within the scope of “clean energy”, we would like to encourage you to apply for the grant. More information on the Foundation can be found on the website: www.tylerlewis.ca
Dr. Fariborz Taghipour Department of Chemical and Biological Engineering, UBC
Prof. Fariborz Taghipour (a CERC member), together with his UBC co-applicants — Prof. Mark MacLachlan (UBC Chemistry) and Prof. Sean Smukler (Faculty of Land and Food Systems) was awarded $2,640,000 grant through the NSERC-SSHRC Sustainable Agriculture Research Initiative (SARI) for his project entitled “Development of a sensor network for sustainable agriculture”.
Project Summary To meet its ambitious climate mitigation targets, it is critical for Canada to invest in sustainable agriculture. The key elements of sustainable agriculture are a) precision farming for efficient nutrient use, which results in reducing greenhouse gas (GHG) production, and b) monitoring GHG emissions, which provides information on the effectiveness of the mitigation actions. In this research project, we will develop new technologies and tools that promote the fundamental elements of sustainable agriculture. We will design, fabricate, and field-test a sensing device to a) monitor nitrogen, a key soil nutrient, to enable precision agriculture, which, in turn, will result in reducing GHG emissions, and b) evaluate GHG emissions through direct measuring, to quantify the impact of the shift to precision nutrient control. Our approach involves applying several emerging technological innovations, including novel nanomaterial synthesize, new sensing platform design, and machine learning techniques, to create arrays of sensing nodes that are activated by UV-LEDs, resulting in a miniaturized and inexpensive multifunctional sensor. The sensor will be used to measure the spatial and temporal variations of nitrogen content and GHG emissions in several agricultural fields, enabling us to evaluate the sensor’s applicability in real-world situations. It will ultimately be possible to deploy our sensors in agricultural fields for real-time mapping of nitrogen and GHG emissions. This would not only streamline informed decision-making regarding the use of fertilizers in each area but also promote sustainable farming practices. Providing high-resolution localized information on plant-available nitrogen and GHG emissions will be critical for agricultural producers to identify emission sources and evaluate mitigation actions when testing and implementing innovations and policies to optimize their operations and reduce emissions. We have developed a comprehensive strategy to translate the research results into application by building a strong coalition at the key levels of academia, government, international research institutions, and industry, with each organization bringing complementary expertise to the project.
