AI-assisted design of photonic components
A multitude of parameters determine the performance of a photonic device, encompassing the optical properties of the constituent materials, structural geometry and dimensions. The availability of simulators to generate data opens the door for the introduction of data-driven modelling, analysis and optimization techniques developed in the field of machine learning and AI in general. Although the use of AI techniques to facilitate photonic design process has been increasing over the last several years, it is still in its infancy. The AI-assisted design of photonics components master project explores AI-based design optimization along a number of directions including design structure, simulation acceleration and accuracy, intelligent search of design space and design fabricability. Appropriate use of AI methods in these areas has significant impact on the outcome of the design process and the time it takes to reach the target design in high-dimensional problems. In addition, it helps domain experts to identify the possibilities and limitations of the design space, opening doors for new research directions.
Master Project Lead
Dr. Yuri Grinberg
Yuri is an associate research officer with the National Research Council and is an expert in applied and theoretical machine learning and reinforcement learning. His primary research interests are in advancing the machine learning state of the art to address problems in physics and engineering. In particular, he is interested in development of appropriate AI methodologies for the design of efficient, small footprint, easy to fabricate photonic components with significantly reduced human effort.
Find out more about Yuri's research interests and his publications.
Contact information
Projects and teams
- AI-assisted miniaturization of integrated photonic components
- AI-assisted inverse design of nanophotonic components
- Accelerated MEMS design optimization enabled by AI with application to optical networks
- Advanced wavefront sensing through machine learning
- Using artificial intelligence for accurate and efficient modelling of complex nanophotonic devices
- AIIR-Power: AI enhanced design and manufacturing of infrared photonic power converters for power and telecom
AI for design of biological systems
AI for design of biological systems master project focusses on AI-based biological therapy design for increased safety, efficacy and specificity with reduced cost and side effects. AI and particularly machine learning is becoming an integral part of modern biological research and is expected to enable more accurate, faster and less expensive innovations in life sciences while at the same time providing prediction of outcomes under different situations. This development relies on implementation of appropriate algorithms, availability of appropriate data and metadata as well as expert input for data labelling, model development and result interpretation. In the long term this master project aims to develop, test and make accessible to non-experts AI-driven simulations combined into digital twins of biological systems and platforms including cellular systems, biodevices and gene editing modalities. These provide methods for unbiased selection and development of optimized cell therapies for different applications.
Master Project Lead
Dr. Miroslava Cuperlovic-Culf
Mira is a Senior Research Officer and Team leader with the National Research Council of Canada. Her work is in the application of machine learning and data mining to life sciences with particular focus of the development of novel diagnostic and treatment methods as well as simulation methods for in silico medicine. Mira has unique training in both experimental and data sciences for molecular and high throughput data analysis and this allows her to work very productively with both experimentalists and clinicians on one side and computer scientists and mathematicians on the other. Within NRC’s AI for Design challenge program she will be working with a team of researchers and collaborators on the development of universal AI methodologies for design of better disease treatments.
Find out more about Dr. Cuperlovic-Culf.
Contact information
Projects and teams
- AI for drug design
- AI for precision discovery on associations in biological systems
- AI for design of multi-targeted therapeutics
- Artificial intelligence protein design for drugs and gene therapies
- AI for simulation of biological systems
- Digital-twin of bioreactor for accelerated design and optimal operations in production of complex biologics
- AI-enabled design of aptamer sensors
- Artificial intelligence powered design of stem cell therapy for degenerative muscle diseases
Deep material science
Improving the material pipeline is important for many societal and technological problems, including energy, high-speed networking and health. AI shows great promise for the design of new materials, design of new processes for materials synthesis, and design of methodologies for the discovery process itself. The deep material science master project leverages a collection of projects to develop AI algorithms and methods that enable material scientists to find better materials. Increased exposure and understanding within the AI community results in more AI practitioners focussing on the many impactful and outstanding problems which exist in the materials domain.
Master Project Lead
Dr. Alain Tchagang
Dr. Alain Tchagang is a research officer with the National Research Council. He has a wealth of research experience combining state-of-the-art methods in statistical signal processing and machine learning to tackle novel and challenging problems in life science, physics and engineering.
Find out more about Dr. Tchagang.
Contact information
Projects and teams
- Automated material synthesis using deep reinforcement learning
- AI for simulation and design of nanocatalytic materials
- Use of artificial intelligence to speed up numerical simulations: application to the design of superconducting tapes for energy and medicine domains
- Fast spectroscopic signatures accelerated by deep learning
- Accelerating design & development of high entropy alloys using machine learning
- Smart microscopy for materials characterization
- AI accelerated mapping & design of CO2-recycling via electrocatalysis at the atomic scale
- Modelling Raman spatial imaging with deep learning
Core AI for Design
AI has the power to accelerate scientific discovery through a combination of big data and intelligent algorithms. Widespread adoption of AI in design however requires advancements in theory, methods and algorithms to address challenges in simulation and modelling and to enable generalizability and adoption by non-AI experts. This master project's research efforts focus on AI methods for increasing the speed of simulations, improving the search of a design space and making explainable models that better align with real-world observation. A key outcome will be AI algorithms that are useful across a very broad set of domains and useable by non-AI experts.
Master Project Lead
Dr. Chris Drummond
Chris is a senior research officer with the National Research Council of Canada and an adjunct professor in the School of Electrical Engineering and Computer Science at the University of Ottawa. Chris's contributions to the AI community are in advancing machine learning research with applications to the monitoring and control of complex systems. The work for which he is best known (with over 1000 citations) is focussed on experimental evaluation, particularly when classes are highly imbalanced, an area of both theoretical and practical significance. Chris participates in the AI community through editorships, chairing conferences and reviewing scientific papers. He is the scientific lead for the core master project within the AI for Design program.
Contact information
Projects and teams
- Phase 1: Designing a friction stir welding process
- Phase 2: Reduced order modelling to speed up simulation of friction stir welding and other design problems
- AI-based shape optimization
- Intelligent Design with Deep Generative Models and Reinforcement Learning
- Quantum Enhanced Design for Materials and Chemistry
- Graph representation learning with limited labeled data for property prediction in design
- On the learning of surrogate models in AI-powered design automation