Innovative Projects Realized

Advances in sensor fusion and gap-filling for satellite based analysis ready surface reflectance data

This project aims to develop a physics-guided deep learning framework that can be used to further enhance the robustness of surface reflectance forecasting in Planet Fusion. We want to improve and refine existing temporal-driven gap-filling techniques in handling extensive cloud cover and dynamic land changes to avoid delays and reduced reliability of delivered insights to the extent possible. The integration of domain knowledge with advanced AI techniques will allow Planet to deliver more robust, continuous, and predictive Earth observation products. This addresses a core need for the company and its clients in agriculture, climate resilience, and land management, who increasingly demand reliable uninterrupted near real-time surface reflectance solutions.

Making Medical AI Smarter, Adapting Language Models for Real-World Healthcare

This international project brings together researchers from Canada and Japan to improve the accuracy and reliability of artificial intelligence (AI) in healthcare. The goal is to enhance how large language models (LLMs)—the technology behind tools like ChatGPT—respond to medical questions by grounding them in verified facts.

At the National Institute of Informatics (NII) in Tokyo, the team is building a Japanese medical language model using real clinical data and a technique called a Mixture-of-Experts. To ensure accuracy, the model will be linked to knowledge graphs, structured databases that help the AI provide more factual and explainable responses. This is especially important in healthcare, where misinformation can have serious consequences. Furthermore, this project will also address how to adapt medical AI systems across different languages and cultures.

The technology developed through this collaboration will directly support and advance the MARVIN chatbot project at Polytechnique Montréal, which helps people living with HIV manage their health. By applying these new tools, MARVIN will become more accurate, culturally aware, and effective for diverse users around the world.

Ergonomic Design of an Automotive Material Sequencing Centre – Year two

The Ford Motor Company is bringing 800 jobs into the Oakville Assembly Plant. These jobs will be concerned with
sequencing parts for the new Material Sequencing Centre. To ensure that workers remain healthy, and their
productivity and quality output is up to Ford’s high standards, Ford (through this fellowship) wants to establish clear
ergonomic guidelines for this type of work. The post-doctoral fellow will conduct surveys in the plant, as well as
review existing ergonomic guidelines within Ford. The main focus of this Elevate award is to develop new guidelines as
necessary for the work in the MSC, and to base these standards off of current research, or research to be conducted
during the duration of this award. The priority is to include strong theory and research to make these guidelines, as
they will be used across all Ford platforms around the world.

American lobster (Homarus americanus) genomics research revitalization

The Canadian Lobster Research Network (CLRN) is a multi-stakeholder and multi-disciplinary research platform led by fishing associations throughout eastern Canada with the principal goal of supporting sustainable and profitable lobster fisheries in the face of rapid climate and ecosystem changes. The CLRN has its origins in the Lobster Node, an industry-driven collaborative research node under the Canadian Fisheries Research Network from 2010-2015. Under the Lobster Node, population genomics approaches were used to delineate the genetic structure of lobster in Atlantic Canada. Weak, albeit highly significant, genetic structure was found at a regional level. A northern and southern population was identified and for the first time, the north–south genetic break was precisely located. Over the last 10 years, genomics techniques have advanced greatly. Current technology would allow for a substantial improvement over previous methods. The previously collected lobster samples are available for re-analysis, but first their quality must be evaluated. This project will inventory and assess the quality of the previously collected lobster samples (4,190) to determine their suitability for additional analysis with updated genomic research techniques, so a lobster genomics research plan can be co-developed with Dr. Scott Pavey (UNB), the CLRN and industry representatives utilizing the samples.

Software Development for Automated Material Property Prediction and Business Intelligence Tools

Chemia is developing an online platform that helps R&D engineers and scientists find, compare, and design better materials for their products using a comperehsnive database of materials, their properties and scientific validation tools. We primarily focus on inorganic materials in areas like clean energy and electronics. This project will enhance the material intelligence, innovation, and integration features of our platform, making them more complete, up to date, and intelligent. The main objective is to automatically add new data from trusted and commercially compliant sources and improve our property prediction tools using AI and first principles calculation tools along with experimentations and experimental data. With a more comprehensive database and integrated validation tools, companies will be able to explore more material options, shorten development cycles, and make better decisions when selecting or designing viable materials.

Vapor Cell Development for Atom-Based Radio Frequency Sensing

Vapor cell technology is the backbone of a wide range of atomic sensors. This project aims to improve the functionality of vapor cells for Rydberg atom radio frequency detection. The main goal of the work is to test vapor cells whose interior walls are coated with materials that passivate the surfaces so as to eliminate the bonding of the sensor atoms to the surfaces, which eliminates the electric field effects that can decrease the sensitivity of the sensor. The interns will make measurements of a collection of vapor cells using optical spectroscopy and x-ray photoelectron spectrsoscopy to optimize the coatings for radio frequency sensing. The statistics of the measurements will be used to characterize the variation of the vapor cell properties and thereby quantify the reproducibility of the fabrication process. The results will be used to improve the fabrication process with the goal of reducing the variation of the vapor cell properties.

Subspace Hierarchies for Musculoskeletal Control

Musculoskeletal systems are responsible for the rich and diverse ranges of motion and behaviors exhibited by all sorts of animals, including humans. Recreating these motions is key for applications in medical simulation, robotics, and virtual reality. Unfortunately, modeling even simple control tasks such as grasping or locomotion with such musculoskeletal systems is extremely difficult. The key problem is that faithfully modeling the dynamics of the musculoskeletal anatomy requires extremely high resolution detailed deformable geometry. This not only makes simulating such systems slow, but it also makes the optimal control task ill-determined. This in turn makes the numerical method used to solve any optimal control problem at hand struggle to converge reliably to a viable solution.
To solve this problem, we draw inspiration from the model reduction community, which simplifies the dynamics of a system by only modeling its most dominant modes. By only modeling a small number of degrees of freedom, and then gradually increasing that number, we aim to construct a reliable optimal control solver for arbitrarily complex musculoskeletal control problems.

Ergonomic Design of an Automotive Material Sequencing Centre

The Ford Motor Company is bringing 800 jobs into the Oakville Assembly Plant. These jobs will be concerned with
sequencing parts for the new Material Sequencing Centre. To ensure that workers remain healthy, and their
productivity and quality output is up to Ford’s high standards, Ford (through this fellowship) wants to establish clear
ergonomic guidelines for this type of work. The post-doctoral fellow will conduct surveys in the plant, as well as
review existing ergonomic guidelines within Ford. The main focus of this Elevate award is to develop new guidelines as
necessary for the work in the MSC, and to base these standards off of current research, or research to be conducted
during the duration of this award. The priority is to include strong theory and research to make these guidelines, as
they will be used across all Ford platforms around the world.

Development of a Climb Model for Aerodynamic Shape Optimization of Blended-Wing-Body Regional Aircraft

In the interest of sustainability, novel aircraft configurations such as the blended-wing-body (BWB) have garnered a great deal of attention for their potential to dramatically improve aircraft fuel efficiency. The BWB concept has been studied extensively using aerodynamic shape optimization to predict its performance benefits and study the key design features of the aircraft. In many of these studies, fuel burn is used as the objective function and is computed using a combination of empirical relationships and a well-known equation for cruise. While this approach is reasonable for long-range aircraft, a more accurate climb model is needed for regional-class aircraft, where a considerable portion of the nominal mission is spent climbing to cruising altitude. This project will address this issue by studying the impact of various climb models on the optimal design and block fuel burn of a regional-class BWB, optimized using a framework developed at the University of Toronto. Existing models will be tested, and novel approaches for computing climb fuel burn will be developed using methods available at Stanford University. The primary result will be the selection of a climb model for use in future studies that will inform industry decisions regarding the next generation of commercial aircraft.

Design and Fabrication of a Pre-Calibrated Mount/Jig and a Fixed Calibration Fixture for WAAM Thermal Camera

Wire Arc Additive Manufacturing (WAAM) is a promising technology for fabricating large metal components, but consistent thermal monitoring is challenging due to camera positioning variability. This project addresses that issue by developing two key systems: a pre-calibrated mounting jig and a fixed calibration fixture.

The partner organization, which operates a WAAM system, requires precise camera alignment to capture thermal data for process optimization and defect detection. The proposed pre-calibrated jig offers an operator-friendly, adjustable mount with linear scales, quick-release mechanisms, and marked focal distances, enabling rapid repositioning while maintaining accuracy. The fixed calibration fixture will attach to the robot arm, ensuring consistent distance and orientation between the thermal camera and wire tip for each setup. Both solutions involve CAD modeling, technical drawing, and fabrication via additive manufacturing, ensuring precision and integration with the existing setup.

The anticipated benefits of this project include improved process repeatability, reduced calibration time, and enhanced data reliability during WAAM monitoring. For the partner organization, this streamlines AM workflows, enhances thermal control, and supports more robust and scalable WAAM solutions. Socially, the project fosters workforce upskilling in digital fabrication, expansion of professional network, and aligning with broader advanced manufacturing goals.

Demand Forecasting Model

Flashana Technologies Inc. develops software products in retail supply chain, specifically predictive analytics for inventory management.

Current challenges include:
(1) Data harmonization, where manual field mapping consumes a large portion of onboarding time.
(2) Algorithmic competitiveness, while competitors deploy classical forecasting, Flashana targets developing AI/ML integrates model that combines customer history with external factors to contribute effective forecasting approaches.
(3) Scenario-based analysis, considering new developments in business analytics, users need rapid AI-guided evaluation of supplier failures or new policies without corrupting production data.

This research project addresses these gaps by building an AI-based model that benchmarks state-of-the-art AI/ML models and time-series demand forecasting methods, and a data-driven simulation environment where users perform what-if scenario analysis and obtain the predicted supply-chain impact.

L2M – ActiMent

Dementia affects over 55 million people globally, with a new case diagnosed every three seconds 1, contributing to annual healthcare costs exceeding $1.3 trillion, including a projected $16.6 billion in Canada by 2031 2. Up to 40% of dementia cases are linked to modifiable lifestyle factors such as sleep and physical activity 3. However, health-conscious individuals lack accessible, scalable tools to proactively manage cognitive health, often relying on costly, invasive clinical assessments like imaging or blood tests that detect issues only after cognitive decline is evident.
The innovation challenge is to develop a practical, data-driven tool that empowers individuals to reduce dementia risk through daily lifestyle adjustments. ActiMent, an Apple Watch app, addresses this challenge by helping user adjust their routines and align with activity patterns associated with better cognitive health. Built on research from over 50,000 participants, ActiMent reveals how specific daily patterns connect to cognitive performance. Unlike fitness trackers or brain games, ActiMent offers actionable, research-driven insights that empower health-conscious users to take meaningful, proactive steps to support long-term brain health. This project aligns with the Lab2Market Validate Health program’s mission to translate health-related research into impactful solutions, moving beyond academic research by validating market potential and developing a commercialization strategy.
Expertise required includes data science and AI to analyze large-scale actigraphy datasets, software development for creating a user-friendly Apple Watch app, and cognitive health research to translate findings into actionable insights. Entrepreneurial skills are also critical to validate market fit and develop a scalable business model through customer discovery and industry mentorship provided by the University of Toronto’s Lab2Market Validate Health cohort.