Projects 2022

Multiple Sclerosis (MS) is the most common neurodegenerative disease affecting young people. Currently, there is no cure. Prof. Arbel is part of an interdisciplinary collaborative research network in Multiple Sclerosis (MS), comprised of a set of researchers from around the world, including neurologists and experts in MS, biostatisticians, medical imaging specialists, and members of the pharmaceutical industry. The team received a Collaborative Network Award by the International Progressive MS Alliance (IPMSA), which has provided Prof. Arbel’s group with access to the first large Magnetic Resonance Image (MRI) MS dataset (~40,000 patients over time) from hospitals around world and from almost all large phase 3 clinical trials for progressive MS. Prof. Arbel’s role in the grant is to develop new deep learning frameworks to automatically discover MRI markers at baseline that are predictive of future patient outcomes (e.g. disability progression, new lesions) in MS, to be used as an outcome measure for use in clinical trials. Recently, her team has developed completely data-driven deep learning frameworks to predict future patient outcomes (e.g. one year ahead) from baseline images, in which latent image features are learned using large amounts of imaging data. The project involves developing explainable deep learning models that explicitly identify the MRI markers at baseline that have a causal relationship with future patient outcomes. This will to permit their clinical interpretation by neurologists and use in clinical trials.

In order to meet the objectives of the project, the student will develop new explainable deep learning models through counterfactual future outcomes. Specifically, given a baseline patient image associated with a future outcome (e.g. the patient will have new and enlarging lesions in a year), the model will provide an alternate (counterfactual) baseline image with the opposite outcome (e.g. no lesions in a year). The differences between these baseline images will permit learning predictive MRI markers associated with future outcomes. The student will work closely with graduate students and Research Assistant in Prof. Arbel’s lab and with members of the collaborating teams, particularly at the Montreal Neurological institute.

Distributed energy resources (DER), which include rooftop solar photovoltaic generation, energy storage systems, and controllable loads like electric vehicle chargers, are re-defining how electric distribution systems are planned and operated. The current modes of planning and operation of these resources are driven primarily by private resource owners' interests (e.g., electricity bill reduction). However, if these resources were to be pooled and adequately coordinated, better system-level outcomes could be achieved (e.g., further decarbonization). In this project, empirical studies will be carried out on test distribution networks to establish how their overall regions of feasible operation are affected by deep DER penetration. Analysis of empirical studies will have to be carried out with the objective of formulating preliminary recommendations for adjusting network planning methods. This project is co-supervised with Prof. G. Joos.

1. Formulate DER scenarios for sample Canadian distribution networks. 2. Carry out power flow studies to investigate how feasibility regions are affected by DER. 3. Analyze results and formulate recommendations for planning studies.

An optical frequency comb (OFC) is a source of light which consists of a series of discrete spectral lines with correlated phase (loosely speaking, a collection of monochromatic waves with the same phase). OFCs have been used extensively in a diverse range of applications, including spectroscopy (studying and using the interaction of light with matter), frequency metrology (involving the precise measurement of frequencies), precision distance measurements, waveform synthesis (e.g., generating arbitrary optical or RF waveforms), optical and RF signal processing, and broadband communications (e.g., optical, radio-over-fiber, microwave, etc.). Important characteristics of an OFC include the number of comb lines, comb spacing (frequency difference between consecutive lines), spectral flatness (variation in power between the different lines), and the mutual coherence of the comb lines (phase relationship). Over the years, a number of approaches for generating OFCs have been investigated, including mode-locking of lasers, exploiting nonlinear optical effects, and electro-optic modulation. Each approach has its advantages and disadvantages in terms of obtaining different comb characteristics. Moreover, the approaches may involve the use of benchtop components and bulk optics or photonic integrated circuits. The objective of this project is to characterize a photonic integrated circuit for OFC generation. The project involves largely experimental work with high speed optical and RF test-and-measurement equipment as well as sophisticated opto-mechanics for coupling light into the photonic integrated circuit.

The photonic integrated circuit involves a subsystem comprising electro-optic modulators in silicon-on-insulator. Working with graduate students, the SURE student will conduct the following experiments: 1. Characterize the performance of the different electro-optic modulators (e.g., in terms of insertion loss, half-wave voltage, bandwidth). 2. Demonstrate OFC generation using the photonic integrated circuit. 3. Optimize the system parameters and settings to obtain the largest number of comb lines possible. 4. [time dependent] Characterize the quality of the OFC, e.g., in terms of measuring the corresponding time-domain waveform and its properties.

In numerous learning tasks, there are strong relationships in the available data, and a graph can often be used to represent these. Recently, there has been an increased research focus on graph neural networks (GNNs) due to their successful application in various learning problems such as node and graph classification, recommendation, and spatio-temporal time-series prediction. In many cases, data becomes available over time, and it is desirable to learn in a continual fashion, updating the parameters of the neural network to adapt to new tasks and data distributions. Ideally, this process should not cause dramatic forgetting of how to perform tasks that were encountered earlier. This project will focus on the development and programming of several continual learning approaches for graph neural networks. The developed algorithms will be tested on applications including recommender systems, time-series prediction and fraud detection.

(1) Literature review (2) Algorithm development (3) Software implementation (4) Data preparation and testing (5) Report preparation

The IMAGE project is aimed at making internet graphics accessible for people who are blind or partially sighted through rich audio and touch. We use rich audio (sonification) together with the sense of touch (haptics) to provide a faster and more nuanced experience of graphics on the web. For example, by using spatial audio, where the user experiences the sound moving around them through their headphones, information about the spatial relationships between various objects in the scene can be quickly conveyed without reading long descriptions. In addition, rather than only passive experiences of listening to audio, we allow the user to actively explore a photograph either by pointing to different portions and hearing about its content or nuance, or use a custom haptic device to literally feel aspects like texture or regions. This will permit interpretation of maps, drawings, diagrams, and photographs, in which the visual experience is replaced with multimodal sensory feedback, rendered in a manner that helps overcome access barriers.

1. investigation and tuning of performance and scaling 2. extending our preprocessors for new types of maps 3. exploring haptic renderings for diagrams and similar textbook materials 4. exploring other commercial ML tools that could be used to extend/improve our existing preprocessors, e.g., scene recognizer 5. porting the Autour server to docker and/or merging OSM and Autour 6. exploring ability to support IMAGE functionality on iOS devices

Most haptics wearables focus on delivery of tactile stimuli to the human body, but rarely consider the face, which is an important area of social touch, especially for couples and parent-children relationships. This project will explore the possibilities for delivering remote touch to the face, and eventually, of feeling a sensation of doing so, in the virtual environment. The architecture will be based on a soft wearable prototype, which operates in conjunction with audio-graphical stimuli in the VR space. We anticipate simulating such interactions as a mother caressing the face of her child, or planting a kiss on the cheek. Applications extend not only to social interaction but further to treatment of medical conditions (e.g., phobia therapy).

Design, implementation, and integration of a prototype within a VR environment, then validate the system through a user study.

In recent years, various single-cell RNA sequencing (scRNA-seq) techniques have emerged. One particular area that these datasets are important is in reconstruction of transcriptional regulatory networks (TRNs). TRNs are directed signed graphs with nodes representing genes and TFs and edges specifying enhancing or inhibiting regulation. Various methods have been developed for reconstruction of TRNs, however due to an absence of gold standard known ground truth for the TRNs, benchmarking these methods is quite challenging. Simulated scRNA-seq data can be helpful in this domain, particularly if the data is generated while taking into account the underlying known TRN. To synthesize single-cell transcriptomic datasets with high biological realism while also considering gene regulation, SERGIO accepts a known TRN. However, this method does not take into account the causal nature of TRNs. Here, we propose a systematic Deep Learning approach to generating single-cell expression data that leverages the causal nature of TRNs dictating TF-gene interactions. In particular, we are interested in causal Generative Adversarial Network (causal GAN) to achieve this goal.

- Implement deep learning models - Apply to real biological data - Evaluate the results - Participate in writing manuscript

Instructors often dedicate a significant amount of time and preparation to promote effective learning. A well-designed course with clear learning objectives and linked assessments, although demanding to accomplish, remains indispensable. However, in conjunction with this comprehensive strategy, it is possible to support successful knowledge acquisition by also incorporating easily adaptable techniques that require minimal effort. These simple techniques, sometimes referred to as “small teaching” [1], do not have to take up more than a few minutes every class and do not require much preparation from the instructor. A good example, taken from [1], is known as ‘prediction’. Using this technique, the instructor invites students to predict the answer to a question. Although they may not have the necessary knowledge to solve it, research has shown that simply attempting to predict the solution increases their learning once they have learned the correct answer. While the literature suggests the potential efficacy of such teaching techniques, their appropriateness in an engineering classroom has yet to be investigated. Engineering education focuses on specific approaches such as problem analysis, design, engineering tools, among others. The goal of this project is to review the existing literature in search of such strategies and identify the ones that will work in the context of engineering education. [1] Lang, James M. “Small teaching: Everyday lessons from the science of learning.” John Wiley & Sons, 2021.

Review existing literature in education and identify simple and easily adaptable teaching techniques that have the potential to work in an engineering classroom.

Noise analysis is a critical part of RF circuits design. In this project we will research noise simulation techniques and implement them in an in-house circuit simulator similar to Spice.

The student will review the literature and identify and build upon state of the art noice simulation algorithms.

In this project we will develop advanced algorithm for analyzing the impact on manufacturing uncertainty on the yield. We will use polynomial chaos based methods in order to obtain the results using less cpu cost than Monte Carlo based methods.

The student will research polynomial chaos based methods, and in particular techniques based on Tensor Train decomposition. An algorithm is then identified and developed for signal integrity applications. The algorithm is then implemented and tested.

The polymerase chain reaction (PCR) is widely used to amplify and identify DNA samples and is now well-known as providing the most sensitive test for COVID19. Most commercial PCR systems require over an hour to produce a result, but we have demonstrated a new approach that uses laser heating of gold nanoparticles to drive the reaction. This has allowed us to produce test results in under ten minutes and opens the technique up to point-of-care applications. Our current system uses fluorescence to quantify DNA amplification; now we are seeking to add more fluorescent channels to allow for more pathogens to be tested for within a single sample. Conventional multichannel quantitative PCR (qPCR) systems use optical interference filters and discrete photodetectors to resolve each fluorescence channel. However, recent advances in CMOS colour sensors and in integrated spectrometers now allow for other approaches. In this project the student will analyse these new options and then implement the most promising into an experimental system. One important consideration will be to find a solution that will ultimately allow for packaging on a pitch of 9mm, to allow for multiple PCR tubes to be tested. The project will be undertaken alongside graduate students who are developing the plasmonic PCR system, and in collaboration with researchers at the Lady Davis Research Institute of the Jewish General Hospital and engineers from a company which is seeking to commercialise the technology. Experiments will be undertaken in the Photonic Biosensing Laboratory in the McConnell Engineering Building.

1. Review and improve existing numerical model for multispectral fluorescence excitation and emission (in Matlab/Mathematica/Python etc) 2. Model the detection response of a CMOS colour sensor and an integrated micro-spectrometer and determine which may be most suitable 3. Procure one of these sensors and test it in the laboratory with different fluorophores

Silicon integrated waveguides are widely used for telecommunications applications and are also of interest for biosensing. We undertake research into several different types of affinity-based photonic biosensors, in which antibodies that are bound to the optical sensor surface are used to capture specific antigens taken from patient samples, thus allowing detection of infection or other medical conditions for point of care applications. One challenge for silicon integrated waveguides as biosensors is that due to the high refractive index of silicon, light is strongly confined to the core and so the evanescent electric field does not strongly interact with the environment. In this project the student will investigate alternative cladding designs for silicon waveguides that could result in increased evanescent wave intensity. This research project will be mainly focused on modelling and analysis, making use of tools such as MATLAB and also photonics modeling software. Full training in the theory and software will be provided and the student will interact regularly with graduate students in the Photonic Biosensing Laboratory.

1. Develop a slab waveguide model for multilayer waveguide for biosensing in MATLAB/Mathematica/Python, considering available cladding materials such as polymers and silicon nitride 2. Optimise the structure to maximise sensitivity to bound analytes 3. Transfer the design to photonic waveguide modeling tool (such as Lumerical) 4. Assess the performance of the design in simulated biosensors and compare with silicon-only structures

In this on-going research project, we consider how to design a broadband wireless access communications system that can adaptably adjust itself to the continuously changing complex environment by using machine learning (ML) techniques. We aim to explore the potential of applying ML techniques to harvest relevant environmental information for improving the resource allocation, performance and operation of the corresponding broadband wireless access communications system. Relevant environmental information can include weather, terrain, propagation (e.g., power, frequencies, etc.), social relationships (e.g., user groups, social networks, etc.) Various ML-based algorithms within a prototype testbed will be developed for the specific topics such as 3D channel modelling/estimation, hybrid ARQ, hybrid massive-MIMO precoding/beamforming, etc., to demonstrate the effectiveness of the Artificial Intelligence (AI)-augmented systems in terms of performance benchmarks such as energy consumption, increase in achievable capacity, reduction in interference, etc. Students will have a chance to understand various new concepts and development tools in both wireless communications (channel modelling, antenna array, beam forming, MIMO, etc.) and machine learning (deep neural network, reinforcement leaning, etc.), and to be involved in practical prototype development, and testing. As an example, one sub-project aims to develop a learning mechanism to jointly optimize Location & signal Power Allocation (PA) in UAV (Unmanned Aerial Vehicle) assisted MU-mMIMO communication system. Other sub-project aims to make use of both the terrain and weather information available in many sources such as Google map, online meteo, to develop a ML-based channel estimator for dynamic resource allocation in a broadband wireless access communication system.

Study the general concepts of ML and wireless communications. Learn how to search for and read scientific papers on a given signal processing or machine learning methods. Investigate Matlab toolboxes, PyTorch, Keras, Tensorflow, and DSP/FPGA hardware for possible applications to algorithm/prototype implementation. Assist in implementation and testing algorithms/prototypes, and in collecting, documenting and commenting the test results. The following skills and experiences are great assets: software development/testing, antenna design, Matlab, Python, VHDL, etc.

Full-Duplex Massive Multi-Input Multi-Output (FD-massive MIMO) Active Antennas Arrays (AAA) are considered for the next-generation broadband communications. Using a massive number of antenna elements can (i) help to adaptively create narrow beams continuously steered to follow the target user while avoiding interference from other users, (ii) increase the communications range, and system capacity. The smart AAA system can follow the mobile user based on (i) a hybrid 2-stage digital and RF precoding structure to reduce the complexity, and (ii) a full-duplex operation for simultaneous transmission/reception over a frequency slot to enhance both spectrum utilization and latency. In this on-going project, we investigate, design and test new promising antenna 3-dimentional structures (such as metamaterials, EBG, dielectric filled, etc.), with integrated power and low noise amplifiers, as well as RF combiners/splitters and smart DSP based control sub-system. The hardware testbed consists of powerful multi-FPGA, multi-microprocessor, and RF Analog-to-Digital Converter (ADC) and Digital-to-Analog Converter (DAC) modules to be programmed with digital signal processing algorithms for transmission and reception/detection of real wireless communications signals.

Study the general concept of Full-Duplex massive MIMO, radio-wave propagation, antenna design and simulation; learn the operation of the antenna design and simulation CAD tools HFSS, Matlab, SystemVue, PCB/DSP/FPGA design tools; learn the operation of RF test equipment Vector Signal Generators, Signal/Spectrum Analyzers, VNAs; prepare the simulation and/or practical test set-ups; assist graduate students and/or research associates to design/test AAA sub-modules, and analyze simulation and/or test results.

Multi-Input Multi-Output (MIMO) has been used in wireless systems such as LTE, WiMAX, Wi-Fi. Full–Duplex massive-MIMO (FD-massive MIMO) is considered for the next generation cellular communications (especially to support broadband Industrial IoT, M2M applications). In this on-going project, we measure and characterize MIMO Self-Interference channels in both microwave bands (for wider signal penetration application) and mmWave bands (for wider bandwidth, high peak data rates applications) and different practical scenarios in order to understand the implications on the FD-MIMO design requirements, in particular, on RF self-interference cancellation. We will investigate various types of channel environments: controlled free-space (e.g., in anechoic chamber), simulated rich scattering (e.g., reverberation chamber), or practical indoor and outdoor environments. Students will have a chance to understand MIMO systems, RF Beamforming based and Baseband Self-Interference Canceller design, Self-Interference and Intended Signal channel measurements, and to work with real-life measurement facilities and testbeds.

Study the general concept of massive-MIMO, radio-wave propagation in free-space and in rich-scattering environments; learn the operation principle of measurement equipment/facilities such as vector network analyzer, spectrum analyzer, vector signal generators, anechoic chamber, reverberation chamber and/or MIMO testbeds; prepare measurement set-ups; assist graduate students and/or research associates to evaluate/analyze measurement/simulation results. Learn how to use Matlab and SystemVue simulation tools.

For power-efficient operation, RF Power Amplifiers (PA) should operate near the saturation region, but this creates non-linear behaviors and distortions in amplified complex signals. Typical approach to balance power efficiency and performance degradation is to use PA linearizers (or PA DPD, PA Digital Pre-Distorters). Non-linear power amplifier with deep memory and very wide operating bandwidth, such as in GaN new power efficient PA technology, is challenging to linearize using conventional models and DPD techniques. In this on-going project we will continue to investigate/test existing state-of-the art practical FPGA/DSP based PA DPD techniques with a single and a network of GaN and GaAs PAs. After the established performance reference, we will develop, simulate and test Deep Neural Network (DNN)-based algorithms to improve and simplify simultaneous linearization of multiple PAs in Active Antenna Arrays (AAA) of massive-MIMO systems.

Study about Power Amplifier characteristics, characterization, measurements and Subsequent Modeling in Matlab. Learn about typical PA parameters such as: gain compression (P1dB compression point), Amplifier Saturation Output Power (P3dB), IMD, IIP3, OIP3 (input/output third-order intercept points), AM/AM and AM/PM distortion. And also about other concepts such as: ACPR (for modulated signals, like QPSK or QAM) and Error vector magnitude (EVM) (for modulated signals, like QPSK or QAM). Help with configuration and interfacing of DPD platforms with a multi-PA AAA m-MIMO testbed, as well as with running the experiments, and pre-analyzing the results under supervision for graduate students. Review literature other conventional and new ML based PA linearization techniques.

Light with its photons allows for data processing. In fact, an optical processor can accelerate the data processing required in deep learning. In parallel, optical devices are experiencing an impressive level of integration leading to exciting new possibilities. Through photonic integration on a chip, the optical processor can be embedded within a modern computer platform used in applications using artificial intelligence (AI). Likewise, similar photonic structures are used in quantum applications. The proposed project is on the development of optical processors for AI and quantum applications. The student must demonstrate excellent communication, resourcefulness, and teamwork skills. We will provide the student with an exciting research environment where we exchange ideas and share knowledge.

Assist in the development of optical processors

This project seeks a motivated student to join a team of experienced graduate students and help develop a software platform using machine learning methods for the design and characterization of next-generation optoelectronic circuits. The qualified student will begin by learning the platform then will assist the team with programming tasks (i.e., testing, debugging, writing, and deploying code), developing usage examples, documentation, and evaluating emerging methods and technologies to identify new opportunities. The qualified student needs to be skilled at software development (preferably in Python or Matlab) and have experience working with numerical optimization methods (e.g., conjugate gradient, Newton, quasi-Newton methods), as well as GPU programming (CUDA) and machine learning methods. An understanding of basic optical and electronic fundamentals, linear algebra, and scientific programming is a strong asset that will accelerate the learning curve required for this project. Lastly, the student will be highly resourceful and technically apt, as the project will evolve with time—using new tools and methods as design objectives change.

Assist the team with programming tasks (i.e., testing, debugging, writing, and deploying code), developing usage examples, documentation, and evaluating emerging methods and technologies to identify new opportunities.

As we integrate more automated systems into our society, whether the systems meet our expectations or not plays an integral role in gaining user trust in intelligent systems. Trust reduces ambiguity by forming expectations of the systems’ intentions and behaviors. Accordingly, people are more willing to adopt automated systems that they trust. However, building human trust toward AI is complex as trust is dynamic, temporal, and incremental. Measuring user trust during human-machine interaction is not feasible today, because trust is a latent variable, which cannot be directly measured. The dynamics of trust between people and that of human-AI are different. Yet, we do not have a working definition of trust in this context. Moreover, we have neither enough nor appropriate data to define trust in the current technological innovation. To build sustainable human-AI trust, conducting longitudinal research is imperative to understand humans’ expectations and reactions towards AI to capture the temporal dynamic of trust. Consequently, the goal of the proposed research is to develop a temporal dynamic model of trust between human and AI by exploring humans’ behavioral patterns and expectations during repeated interactions with an AI system. More specifically, the study will be conducted at the context of RIL at McGill, where naturalistic and realistic human-AI system interactions occur.

Conduct ethnographic studies (user diary, interviews, observations) at RIL to understand shopping behaviours of diverse users at RIL to discover at which point users’ mistrust occurs and how the mistrusts are recovered. Based on the analysis, journey map of longitudinal trust will be mapped to define when or how interventions could occur to gain users’ trust of the AI-based service embedded system.

Low-power microwave breast cancer detection is based on the reported inherent dielectric contrast between healthy and cancerous tissues. This underlying hypothesis promises tumor detection in a safe manner, and very different from the currently used modalities (X ray mammography, MRI, ultrasound). Our group uses a radar-like approach while striving to detect embedded tumor non-invasively and through frequent screenings. The ultimate goal is to have a system that is wearable and allows for frequent screenings in small clinics and even at home, so that the tumors are detected at the early stage, thus increasing the chance of successful treatment. We are currently working on the improvement of the system prototype, so the student working with our team will be exposed to all system components and the research towards their possible improvement: antennas, switching matrix, control circuitry, signal processing and the construction of phantoms for experiments in a controlled laboratory environment.

The student will be first introduced to the overall system functioning. Thereafter, based on the stage of the project, the student will be assigned to work in-depth on a concrete system component. For example, this can be a new antenna design, the optimal number of antennas and their layout. The student is expected to attend regular research group meetings.

Skin cancer is usually detected visually, and the samples from the subsequent biopsy of a suspicious lesion are sent for pathological analysis. Literature reports on different dielectric properties between skin tumors and the healthy skin in the microwave- and millimeter-range. Hence, sensors developed for these frequencies have the potential to non-invasively detect malign tumors. As a starting step, we need to develop a range of reliable phantom skin models which would allow for controlled laboratory testing. These models serve for parametric design studies, meaning, for constructing simulated skin-tumor combinations. These can be measured to validate the approach and proceed with human studies. The ultimate goal is a simple, inexpensive tool, which an aid the dermatologist in the diagnostic process.

Literature search and reading of relevant papers, experiment design, skin/tumor phantom construction, possibly use of dielectric probe for characterization. The student is expected to attend group meetings.

Alkali intercalation of graphite is critical to lithium ion battery operation. It is also an active area of research for its potential to dramatically modify multi-layer graphene and graphite physical properties. Most striking is the Lifshitz transtion, which entails a modification of Fermi surface geometry and transition from n-type to p-type behaviour. Previously observed in photo-emission studies, charge transport measurements of multi-layer graphene and graphite undergoing a Lifshitz transition have never before been measured. The goal of this project is to contribute to an on-going experimental program to measure, in real time, the charge transport characteristics of multi-layer graphene and graphite to contribute to understanding of alkali intercalation and observe the Lifshitz transition.

Student will work with a senior graduate student. Reesarch tasks include exfoliation of multi-layer graphene and graphite on quartz, fabricate metal electrodes in Hall bar geometry, and package by flip-chip methods with a miniature alkali vapour source in a glove box. Using ac Hall magnetometry, measure the charge transport characteristics including kinetics of alkali doping. Previous experience with exfoliation of 2D materials is desireable.

Waste electrical and electronic equipment (WEEE) is the fastest growing solid waste stream globally, exceeding 50 megatons per annum. Precious metals enter landfill, and when metal reclamation is performed, it typically involves toxic discharge and unregulated labour in jurisdictions where enforcement is challenging. This project concerns the search for new circuit substrate materials to replace existing fiber glass technology, focussing on the graphene oxide material family. Graphene oxide is soluble in solvents, facilitating metal and component recovery at circuit end-of-life. The specific goal of this project is to assist in the development of graphene oxide circuit substrate prototypes using pre-existing formulations of cross-linked graphene oxide. Previous experience with circuit board layout and assembly is required.

Working under the direction of a senior graduate student, the student will design prototype circuit demonstrators for implementation on semi-rigid and rigid graphene oxide substrates. Student will metalize substrates by shadow mask methods, and assmeble circuits using low-temperature bonding methods. Demonstration of circuit functionality is essential.

Conventional design approaches, such as system-on-chip (SoC), suffer from scalability and heterogeneity limitations. Chiplet-based integration platforms, such as Si-IF, FlexTrateTM, and interposers, offer promising solutions to overcome these bottlenecks. Thermal management is a critical challenge for realizing state-of-the-art chiplet-based integration platforms, especially for high-performance applications. Therefore, an easy-to-use and computationally efficient thermal simulator that supports chiplet-based platforms is desired. In this project, the student will adapt an existing thermal simulator to accommodate heterogeneous integration platforms. Qualifications: • Advanced C++ programming skills, including object-oriented programming, writing scripts, and debugging [required] • Familiarity with shell scripting and other programming languages, such as C, JAVA, Python, and Perl. [preferred] • Familiarity with basics of heat transfer concepts [preferred]

There exist several compact simulators such as HotSpot, which support thermal simulations for SoCs and three-dimensional integrated circuits (3D ICs). The objective of this research project is to modify and add features to the available compact simulators to extend their support for chiplet-based platforms. The code of the latest version of HotSpot (V 7.0) will be used as the base. The upgraded platform should be able to support compact thermal simulation for three targeted chiplet-based integration platforms, including Si-IF, FlexTrate, and interposers.

With the growth of the Internet, the amount of information available to research teams has been increasing exponentially in the past two decades, which results in significant information overload. Approaches for manual knowledge extraction and curation does not scale up in practice. The project aims to take technical documentation such scientific papers, source code repositories, videos, blog posts, or patent documents as input to extract technical concepts and key relations between those concepts in the form of a knowledge graph by using various machine learning techniques. The project aims to provide an extensible framework to incorporate future concepts and relations. The project likely involves a close collaboration with an industrial partner.

* Review relevant literature (machine learning, knowledge graphs) * Extraction of technical concepts, creators * Mining of relations between concept, resources and creators * Develop adaptors to various data sources * Carry out experimental evaluation using benchmarks

While automated decision making assisted by various machine learning techniques has had significant success in online retailers with a massive number of online customers, the take-up of such initiatives is much slower in traditional in-store retailers where decision making is still often dominated by the judgement of retail experts. Multi-armed bandit algorithms are reinforcement learning techniques which help optimize decision making in an uncertain environment with a wide range of applications. The presence of delays, i.e., when subsequent decisions need to be made before all feedback from previous decisions is known, represents a particularly challenging subclass of decision-making problems. The main goal of the project is to evaluate the use of multi-armed bandit algorithms for price optimization in retail applications and provide adaptations for such retail scenarios, if necessary. Given historical data on price variability or detailed time series data on how the prices of products change over time, multi-armed bandit algorithms will help provide a recommended price for a given product. The project will be co-supervised with Prof. Maxime Cohen (Professor; Co-Director of Retail Innovation Lab at Bensadoun School of Retail Management) and it is planned to involve collaboration with an industrial partner.

- Literature review (multi-armed bandit algorithms, A/B testing) - Exploratory data analysis of historical data of prices - Develop, integrate and execute bandit algorithms on various datasets - Performance evaluation and benchmarking of bandit algorithms on such datasets

In this project, a team of students will investigate the ways to augment the design process with the automated verification focused on ensuring security and trust. We will consider the systems based on FPGA substrate, as well as the RISC-V processor core, to which the secure applications would need to be ported. The incorporation of Intellectual Property (IP) Cores for authentication, encryption and related tasks will be the common mode of operation. Higher-level applications of secure systems in Internet of Things (IoT) will be considered as well. The project will also rely on the use of the state-of-art assertion-based verification tools, such as MBAC.

Incorporation of the design procedures into larger flows Case studies, verification engineering Validation and benchmarking

The Aerospace Mechatronics Laboratory houses a wide range of unmanned aerial vehicles, including airships, quadrotors, gliders, fixed-wing and hybrid aircraft. The overall objective of our research is to develop platforms for a range of tasks. Example applications include gliders for wildfire monitoring and fixed-wing aircraft for autonomous acrobatic flight through obstacle fields. Two SURE students are sought with strong interest and aptitude for research in the areas of robotics, mechatronics and aerial systems. One of the positions will be oriented toward design and modeling of small indoor airships. The second position will focus more on the modeling and control of fixed-wing UAVs. Some experimental testing of components and associated flight tests will be involved, particularly for the second position. In addition, the students will be involved with interfacing new sensors into the platforms, for the purposes of acquiring data and for closed loop control. Some programming experience would be useful for the development of a real-time hardware-in-the-loop simulation. The students are expected to assist with hardware interfacing, programming, conducting experiments, and processing the data.

The tasks will be varied and could accommodate mechanical, electrical or software engineering students; but ideally someone with experience in all aspects. The first position will be best served by a student with knowledge of CAD modeling and Matlab/Simulink and/or other physics-based modeling tools such as Gazebo and ROS. The second position will be best served by a student with knowledge of interfacing of sensing hardware with microprocessors; and programming. Both students will be involved in experimental testing in the field.