Professor Mark Jeffrey (EngSci 0T9, ECE MASc 1T1) is one of four new faculty members joining The Edward S. Rogers Sr. Department of Electrical & Computer Engineering (ECE) this summer. Jeffrey received his PhD degree from the Massachusetts Institute of Technology in 2019. His research is in the areas of computer architecture and computer systems, with an emphasis on parallel computer architecture, parallel programming models and hardware/software co-design for parallelization, performance and efficiency. Writer Jessica MacInnis spoke to Jeffrey about his research, why joining ECE at U of T excites him, and his favourite way to get to know a new city.
Tell us about your research.
I work on computer architecture, spanning computer hardware, software and the interface between them.
Computer users and programmers always have (and always will) want better performance from their computer systems, including faster running time or less drain on battery power. However, the technology trends that enabled decades of exponential performance improvement are screeching to a halt. To sustain future performance scaling, we need radical new architectures that more efficiently use existing resources (transistors) on a chip so that we can continue to solve ever-growing computational problems and support new applications.
In particular, parallelism — running many computations at the same time — is an old yet vital tool to improve performance efficiently, but in many applications it is still hard to wield. For example, you and a friend could prepare a single meal together twice as fast as you would alone. But could four friends prepare the meal four times as fast? What about 100 friends? Without extremely careful orchestration, there are “too many cooks in the kitchen,” which slows down the process because they are sharing one kitchen, i.e., changing shared state.
The same is true in parallel computer architectures and domain-specific accelerators. It is straightforward to break up some applications into parallel tasks, but for the majority of applications it is too difficult, or even impossible, to express enough parallelism on current architectures.
What excites you about this research?
I am excited to rethink the role of hardware, software and their interface, developing new ways to continue energy-efficient performance growth.
Having come to architecture from a software background, I empathize with programmers, which drives me to flexible solutions that keep them happy and productive. One of my approaches is to study application domains and algorithms to identify performance bottlenecks, reconsider which bottlenecks are fundamental and which are artifacts of current interfaces, and then develop new programming models and hardware architectures to solve them.
This applications-to-microarchitecture approach gives my students and me the opportunity to learn from and collaborate with other experts who work at many levels of the hardware-software stack: applications, compilers, architecture and hardware design.
Why did you choose ECE at U of T Engineering?
I chose ECE at U of T for the excellence of the faculty and the talent and drive of the graduate and undergraduate students. I am thrilled to join my colleagues in the Computer Engineering Group with deep expertise spanning applications, computer systems, hardware design and my home of computer architecture. The department attracts phenomenal students from across Canada and the world, providing a diverse environment for education and research to thrive.
Also, Toronto is a great place to live!
Any collaborations or interdisciplinary work you are most looking forward to pursuing?
U of T is known internationally for its high-quality researchers. I cannot wait to collaborate with domain experts within ECE and across faculties to unleash compute performance in applications with untapped opportunities. These opportunities could include machine learning graphical models, electronic CAD, robotics path planning, mathematical optimization and more.
Any advice for the incoming ECE class?
Have faith in yourself, never stop learning, and keep an open mind. It might take time to find the topics that interest and drive you. You might not have heard of them yet. You have time, and the wait is worth it.
Tell us a fun fact about yourself.
I love music and am an amateur drummer and percussionist. I have had the opportunity to learn and work in many cities, and joining an orchestra or jazz band has been my favourite way to get to know a place and (some of) its people.
Two U of T Engineering professors have been awarded new Canada Research Chairs.
Professor Gisele Azimi (ChemE, MSE) now holds the Canada Research Chair in Urban Mining Innovation, while Professor Arthur Chan (ChemE) is the new Canada Research Chair in Atmospheric Chemistry and Health.
Azimi’s research focuses on innovative ways to recover valuable elements from unconventional sources. These include rare earth elements, such as neodymium and dysprosium, which are needed for the types of batteries and magnets found in technologically advanced products such as smartphones, electric vehicles and wind turbines.
Using unconventional techniques such as supercritical fluid extraction, Azimi and her team are developing new ways to recover these valuable elements from materials that would otherwise be discarded as waste — old electronic devices, batteries left over from electric vehicles, and even byproducts of the aluminum industry.
By converting waste into valuable products and developing extraction methods that are use less energy or produce fewer byproducts, Azimi is at the leading edge of a more sustainable resources sector. Earlier this year, she earned the Faculty’s McCharles Prize for Early Career Research Distinction.
Arthur Chan and his team study air quality, particularly the impact of particulate matter such as organic hydrocarbons or heavy metals on human health. These compounds can be carcinogenic or have other toxic effects. The goal is to measure these pollutants in urban air and trace their sources in order to prevent future pollution.
Recently, the team collected dust from homes in Fort McMurray, Alta., analyzing it for evidence of harmful toxic substances left in the aftermath of the devastating 2016 wildfire. Their study revealed normal levels of contaminants that are comparable to homes across Canada, and so far, no evidence of long-term health risks from fire-ash exposure in residents’ homes.
“These new Canada Research Chairs will accelerate the work of top researchers in critical areas and help translate their innovations into new technologies, processes, and business models,” said Ramin Farnood, Vice-Dean, Research at U of T Engineering. “We are very proud of the work that Professors Azimi and Chan are doing to build a more sustainable world.”
Professor Philip Asare joined U of T Engineering’s Institute for Studies in Transdisciplinary Engineering Education and Practice (ISTEP) and the Division of Engineering Science (EngSci) this month. Previously, he was Assistant Professor in Electrical and Computer Engineering at Bucknell University, and before that spent time as a scholar-in residence at the U.S. Food and Drug Administration while earning his PhD in computer engineering from the University of Virginia.
Writer Tyler Irving talked with Asare to find out more about his research and his philosophy of engineering education.
Can you talk about your background in STEM? Why did you choose engineering?
I decided to pursue engineering in the hope of using my skills to better the human condition. This is why I gravitated towards medical applications in my work, such as assessing and assuring safety of emerging medical devices.
Along the way, I realized that in North America, many people who are involved in engineering, and in technical fields more broadly, did not look like me or share a similar background. I grew up in Ghana, where my mom is an engineer, and most of the engineers I knew were like me (though we still have problems related to gender and socio-economic status).
Being aware of the underrepresentation led me to get involved in outreach efforts, even as an undergraduate. I’ve been on a quest to address issues of diversity, equity, and inclusion in these fields since.
I decided to get into higher education because I felt that in addition to the impact I could make from my own work, I can also help train the next generation of science, technology, engineering and mathematics (STEM) professionals to be attentive to issues of the human condition including social justice, equity and inclusion.
You’ve said that you see engineering as a “humanist enterprise.” What does that mean?
Engineering is a human activity geared towards human goals, full or rich and interesting stories of how particular ideas or products came to be. The human dynamics involved in the process have a significant effect on the outcome — who is doing the engineering matters a lot.
There’s a quote in the book Plato and the Nerd by Edward A. Lee of UC Berkeley, that I love:
Technology is not a collection of Platonic truths that have always been lurking in the background, waiting to be discovered, but is rather a rich sociological tapestry of ideas created by human inventors. It is shaped by those humans, and had a different set of humans created it, including more women, for example, the technology would unquestionably be different.
Why did you choose U of T Engineering?
I loved the idea of ISTEP and the goals they had in mind when it was set up. The same is true of the Engineering Science program. In my career, I have looked at engineering often from a design perspective, so the fact that this position would be focused on design and housed in both ISTEP and EngSci really resonated with me.
I feel like the question of “how should we prepare the next generation of engineering leaders for the world they will face in 2050?” is something I’ve been asking myself for a while, and the opportunity to explore this with folks here was really attractive.
In addition to that, Toronto works out well for my family, and U of T’s size, diversity of research fields, and prominence mean that I get to interact with a lot of different people, which is exciting for me.
What are your plans for teaching and research?
The overarching theme will be “(re)humanizing design”. Course-wise, I’m responsible for the second-year Engineering Science design course Praxis III as well as the capstone course for EngSci electrical and computer engineering majors.
On the research side, I’m going to continue my work on increasing representation of underrepresented groups in STEM. I also have some work on how we educate engineers to operate in the complex socio-technical world we have created using various engaging pedagogies that will also be continuing. In addition, I will be continuing my work on system design, especially in the medical area.
As a new professor, what one piece of advice would you give to new students?
Learning happens in many places, and not just in the classroom or in a course. Pay attention to those learning moments that happen outside traditional spaces. Take time to get to know people deeply, especially people who have different backgrounds and experiences than you do.
University is one of the few times in life you’ll have your time mainly focused on learning and making connections with others. Make the most of it.
Exposure notification apps, such as COVID Alert — to be launched by the Canadian federal government this month — could be key tools for countries as they work to reduce the spread of COVID-19.
But as public health officials are evaluating COVID Alert’s efficacy in the fight to stop the spread of the virus, Professor David Lie (ECE) is closely watching the app to see just how closely the app is watching you.
Lie is part of a team of researchers who recently published a paper exploring the constitutional implications and technological underpinning of apps, such as COVID Alert in Canada. The study is a collaboration between Lie and a team of legal scholars.
“With apps like this, there’s a spectrum that ranges from very useful as a method of contact tracing, to extremely invasive in terms of your privacy,” says Lie. “There’s always a trade-off, and what we found was that the COVID Alert app leans heavily on the side of privacy protection, which is great and should allay many of the fears people have about app-based exposure-risk tracking, but at the same time, it means that its potential as a health tool is not fully realized.”
Contact tracing has long been used to track and notify people possibly infected with a virus or disease. The process is a labour-intensive public health endeavour that involves identifying people who have been exposed to a virus and mapping out their contact with others after that individual was infected, then contacting those people and doing the same.
Exposure notification apps are complementary to traditional contact tracing in that they can reduce the time and alleviate the human power required to identify and notify contact between people by having your smartphone do the tracing. In the case of the COVID Alert app, proximity data, via Bluetooth, is used. As people carry around their smartphones, a randomized signal is sent from one smartphone to another in close range.
This proximity data — the randomized “messages” sent from phone to phone — are uploaded to a server. The app then pulls these messages from the server to compute a risk score based on if, and how long, it has come into the contact with the phones of individuals who have tested positive for COVID-19. The app would then alert people who had high scores to get tested.
“Although the app tells you that your phone has come close to another phone carried by someone who has tested positive, it can’t tell you at what time, or where that contact occurred,” explains Lie. “Bluetooth doesn’t know whether you were standing in an elevator with that person or if you were sitting in separate cars at a red light next to each other.”
The paper, which reviews the potential benefits of contact tracing and exposure notification apps and the limitations of the technology, was co-authored by Lie and several legal experts, including Professor Lisa Austin of the Faculty of Law and cross-appointed to ECE.
“Exposure notification apps could have an important role to play in responding to the COVID-19 pandemic here in Canada,” says Austin. “The Canadian Charter provides an important framework for how to balance rights — like privacy rights — in a free and democratic society, but it’s incredibly important that as we understand the legal framework and implications of these apps, we also understand the technology so that we can properly evaluate and review the benefits and consequences.”
The technology behind the COVID Alert app is anonymous and decentralized, meaning that no company or authority can get to all the data and the data is held by the individual who can choose when that information is released.
“Making the app a more effective contact tracing tool would involve using readily available technologies such as GPS-location data tracking or a centralized registration system — but this comes at a cost of privacy,” says Lie.
“But there are other things to keep in mind too, like how many people download and use the app, it’s not clear what the threshold will have to be to make any app as effective as possible — unless a government mandates its use, which would be another question for legal experts,” adds Lie, who also raises the question of fairness, as many Canadians do not have access to a smartphone or reliable cell phone reception.
An individual’s comfort with using an app, the privacy trade-offs of tracking a person’s movement, combined with the technological limitations and possibilities of a smartphone are all considerations that Lie and this transdisciplinary team are looking at.
“This collaboration between researchers in the Faculty of Law and ECE is just one example of the importance of transdisciplinary work to solve these really complex issues we are facing,” says Professor Deepa Kundur, Chair of ECE. “We can no longer look at the implications of technical advances separately from other issues — by working with other disciplines instead of beside them, electrical and computer engineers are able to make greater contributions towards solving big challenges.”
“I think this is a very real and urgent example of how we have to look at transparency and privacy in the 21st century,” says Lie. “We can’t truly understand the legal ramifications in our digital world unless we understand the technology and vice versa, and in this case, we are looking through these lenses during a global pandemic with very serious health and economic ramifications.”
Researchers from U of T Engineering have used the Canadian Light Source (CLS) at the University of Saskatchewan to improve their techniques for converting CO2 into ethanol, a valuable chemical that can be used both as a fuel and in a variety of industrial applications.
Ethanol produces fewer emissions when compared to gasoline, but the renewable fuel is most often made from corn and wheat so there is a strong interest in non-food production methods. By capturing and converting carbon emissions to ethanol, the fuel’s environmental benefits could be multiplied.
The research team led by Professor Ted Sargent (ECE) focused on producing chemicals through CO2 conversion—such as ethanol, ethylene and methane—helping to transform harmful greenhouse gases into useful products. The group aims to produce the target chemicals, in this case ethanol, with high outputs and minimal energy inputs.
The ethanol project is sponsored by the Canadian energy company Suncor, which invests in cleaner, renewable fuel sources, the Natural Sciences and Engineering Research Council (NSERC) of Canada, and the CIFAR Bio-Inspired Solar Energy program. If successful, more Canadian renewable fuel like ethanol could be created from greenhouse gases and less from farmed food.
In a recent paper published in Nature Energy, the team studied ethanol production by developing catalysts through overcoating copper with nitrogen-doped carbon (N-C/Cu). To understand how the catalysts work and provide valuable information for fine-tuning its efficiency, they studied the structure and chemistry of catalysts using the SXRMB beamline at the CLS.
They discovered that gaps below 1 nm between the Cu and N-C layers may act as a reactor during CO2 to ethanol conversion, a property that could be harnessed to maximize output.
“We also found that the nitrogen-doped carbon layer is very important for producing ethanol because it can promote the reaction’s selectivity to ethanol,” said Xue Wang, a postdoctoral fellow in Sargent’s group.
Wang explained that ethylene and ethanol, two valuable chemicals that can be created from CO2, are derived from a shared key intermediate (HOCCH*) in the conversion reaction, but the end result depends on whether C-O bond-breaks from HOCCH*, which results in ethylene, or stay stable, which results in the desired product ethanol.
“Our goal is to suppress the C-O bond breaking from HOCCH* so less ethylene will be produced. We found that this catalyst is very efficient for creating ethanol,” Wang said.
The researchers also discovered that the N-C/Cu catalyst is a strong and stable system, which is necessary for potential commercialization. In their experiments, they achieved 15 hours of stability at 52 percent Faradaic efficiency (FE) with a full cell energy efficiency (EE) of 16% in creating ethanol. The researchers aim to further improve the selectivity to ethanol, production rate, EE, concentration of ethanol and operational stability in order for the system to be used successfully in industrial applications.
This story originally appeared on the Canadian Light Source website.
Professor Shurui Zhou is one of four new faculty members joining The Edward S. Rogers Sr. Department of Electrical & Computer Engineering (ECE) this summer. Professor Zhou received her PhD in the Institute for Software Research at Carnegie Mellon University’s School of Computer Science.
Her research focuses on helping distributed and interdisciplinary software teams to collaborate more efficiently, especially in the context of modern open-source collaboration forms, fork-based development and interdisciplinary teams when building AI-enabled systems or scientific software. Writer Jessica MacInnis sat down with Professor Zhou to ask her about her research, what drew her to ECE at U of T and how her research led to a collection of utensils.
Tell us about your research.
Software development requires groups of people to work collaboratively. Nowadays, these software teams have become increasingly distributed: there are people working from home and even teams spread across continents, in different time zones. In open source communities, there are people who might not even know each other, but are nevertheless building software together.
In addition, the development process requires stakeholders with different backgrounds to collaborate together, from requirement analysis to software design to software delivery. The collaboration efficiency has a direct impact on the quality of the outcomes.
The vision of my research is to help distributed and interdisciplinary software teams to collaborate more efficiently. I combine advances in tooling and software engineering principles with insights from other disciplines, such as organizational theory and social computing.
I mix a wide range of research methods to understand the problem better. Throughout my work, I often start with studying a problem empirically, such as conducting surveys and interviews with stakeholders, before applying theoretical tools from other fields. I use all of that knowledge to design new tools or software engineering practices, and to evaluate the solutions empirically and pragmatically.
What excites you about this research?
What excites me is the practical strategies that can come out of this work. I love helping software teams in industry resolve real problems in a systematic way.
Any collaborations or interdisciplinary work you are most looking forward to pursuing?
I am excited to collaborate with many colleagues at U of T, both within and outside of the ECE department. For instance, I am interested in helping AI experts work effectively with software engineers. I would like to collaborate with Professor Nicolas Papernot to understand the challenges of building AI-based software systems, and with Professor Alison Olechowski on building bridges between software and hardware.
Any advice for the incoming ECE class?
I would encourage students to strengthen their interdisciplinary thinking, which is becoming a highly necessary and frequently sought skill in the world. And, have fun!
What do you hope to accomplish, as an educator and as a researcher, in the next few years?
I would like to build a successful software engineering research group and attract students to do high-impact research.
I would also like to help the next generation of engineering students become software development and educational leaders who will solve the problems associated with building large-scale and critical software systems.
Tell us a fun fact about yourself.
I collect forks as I have spent six years of my PhD studying how to help people use forks to build better software systems!