EmergeNeed Wins Treehacks 2022: How an ECE Master's Student is Shaping the Future of Medical Diagnosis

EmergeNeed Wins Treehacks 2022: How an ECE Master’s Student is Shaping the Future of Medical Diagnosis

Shreya Verma is a Master of Applied Science student in the UBC Department of Electrical and Computer Engineering, who has worked with the UBC Faculty of Medicine and Vancouver General Hospital to develop algorithms to detect skin cancer in its early stages. After being diagnosed herself with a potentially fatal tumor, and her recovery from her diagnosis, Shreya was inspired to pursue a research career in the field of early medical diagnosis. Shreya is also passionate about biomedical AI and is currently developing a deep learning algorithm to detect sepsis in infants in the Neonatal Intensive Care Unit. As part of the team, Shreya’s project EmergeNeed has won the Treehacks 2022 Covid Challenge and Sponsor’s Challenge, which we were excited to connect with her about to learn more about her story.

Read on to learn more about EmergeNeed, Shreya’s experience in our ECE department, and some great advice from her for any students currently looking for help or facing difficulties or mental health concerns in their programs.

Can you tell me a bit about EmergeNeed?

EmergeNeed addresses three main areas of healthcare improvement. To begin, there is no easy way for an individual to know how crowded a hospital will be at a given time. Especially in the current pandemic environment, users would benefit from information such as crowd level and estimated travel times to different hospitals near them. Knowing this information would help them avoid unnecessary crowds and the risk of COVID19 exposure, receiving faster medical attention and an enhanced treatment experience. Additionally, such a system allows hospital staff to operate more effectively and begin triaging earlier since they will receive a heads-up about incoming (non-ambulance) patients before they arrive.

Moreover, online information is often unreliable, and specific demographics may not have access to a primary care provider to ask for advice during an emergency. Our interface allows users to access on-call tele-network services specific to their symptoms easily and therefore receive advice about options such as monitoring at home, urgent care, or an emergency hospital.

Finally, not knowing what to expect contributes to the elevated stress levels surrounding an emergency. Having an app service that encourages users to actively engage in health monitoring and providing tips about what to expect and how to prepare in an emergency will make users better equipped to handle these situations when they occur. Our dashboard offers tools such as a check-in journal to log their mood gratitudes and vent about frustrations. The entries are sent for sentiment analysis to help monitor mental states and offer support. Additionally, the dashboard allows providers to assign goals to patients and monitor progress. For example, taking antibiotics every day for 1 week or not smoking. Furthermore, the user can track upcoming medical appointments and access key medical data quickly such as their COVID19 vaccination card, immunization forms, and health insurance.

What inspired you to create EmergeNeed?

“Emergency” + “Need” = “EmergeNeed”

Imagine a pleasant warm Autumn evening, and you are all ready to have Thanksgiving dinner with your family. You are having a lovely time, but suddenly you notice a batch of red welts, swollen lips, and an itchy throat. Worried and scared, you rush to the hospital just to realize that you will have to wait for another three hours to see a doctor due to the excess crowd. This is exactly what inspired us to create EmergeNeed.

Now imagine that you could quickly talk to a medical professional who could recommend going to urgent care instead to treat your allergic reaction. Or, if you were recommended to seek emergency hospital care, you could see the estimated wait times at different hospitals before you left. Such a system would allow you to get advice from a medical professional quickly, save time waiting for treatment, and decrease your risk of COVID exposure by allowing you to avoid large crowds

How do you feel about having EmergeNeed win Treehacks’ 2022 Covid Challenge and Sponsor’s Challenge?

I had applied to be a part of TreeHacks 3 years ago and I was not selected. So this year to be not only selected but also to win the challenge feels surreal! Sounds cliché, but I am glad I didn’t give up. From team building to presenting my project to 50+ people, it was a great experience. It was a bummer that I could not attend it in-person, but I definitely will next year! A friend of mine told me during grad school small wins are important and this win came at the time I needed it the most.

Why did you choose to study at ECE?

I was diagnosed with a potentially fatal tumor which majorly impacted me physically and mentally. However, it also led me to the brilliant acknowledgment that timely and good quality treatment had saved my life. It was at that moment that I decided to study and explore the field concerning early diagnosis of medical irregularities,playing my role as a mindful engineer by contributing to saving lives. This led to me to apply for the Visiting International Research Student position offered by UBC. I worked with the Faculty of Medicine and Vancouver General Hospital to develop algorithms to detect skin cancer in its early stages. My time here made me realize that there is no better place than the ECE department at UBC to work towards my goals.

How does your career path and graduate experience connect to your current position?

I was always fascinated with Biomedical AI. Currently, I am currently a part of Dr. Guy Dumont’s research team at BC Children’s Hospital. My team and I are working on the estimation of respiratory rate and heart rate of infants in the Neonatal Intensive Care unit using RGBD data. I have learnt alot about my field by taking exceptional courses offered by professors at UBC, such as Advanced Machine Learning, Trustworthy Machine Learning, and Deep Learning in Digital Media, to name a few.

What advice do you have for students who may be struggling or feel as though they are not doing enough?

Just three words – Hang in there. I started my Masters during COVID in September 2020. I didn’t know anybody and was too shy to reach out to other students over Zoom. I felt I was not doing well in terms of academics, research, basically just letting everyone down. On top of that, living in such an expensive place as Vancouver is stressful, especially when you are an international student. I am sharing this so that I can let my peers or even incoming grad students know that it is okay to feel that you are not enough and that everyone goes through this phase. I have always relied on UBC Applied Science counseling services when I feel overwhelmed (https://students.engineering.ubc.ca/health-and-wellness/). Another thing that helped me is being a part of the ECEGSA. I met so many fellow grad students and also organized fun activities like skiing at Grouse mountain, coffee socials, and a board game night! I truly believe “One day you will tell your story of how you overcame what you went through and it will be someone else’s survival guide” and that is what I am striving for.

What are you looking forward to in the future?

I am right now working towards finishing up my Masters. With that I am also preparing for job interviews (so A LOT of Leetcode). I am very inspired by a fellow graduate student that made it to Forbes 30 under 30, and that is what I am working towards!

What are some of your favorite activities to help you unwind while you’re working or studying?

I am an outdoor person! I am out hiking or skiing every other weekend! I love playing basketball, so I am also found on the court very often.

ECE Professor wins 2022 IEEE Cyril Veinott Electromechanical Energy Conversion Award!

Dr. Juri Jatskevich, a UBC professor in the Department of Electrical and Computer Engineering, has won the 2022 IEEE PES Cyril Veinott Electromechanical Energy Conversion Award! This award is given to those who have made outstanding contributions to the field of electromechanical energy conversion, and is administered by the IEEE PES Awards Committee. Dr Jatskevich is recognized for his contributions to electrical machine models for transient simulations of electromechanical energy conversion in power systems.

Dr. Jatskevich received his Bachelor’s in Electrical Engineering from Ukrainian National Agricultural University, and completed both his Master’s and Doctorate at Purdue University. He has chaired the IEEE CAS Power Systems & Power Electronic Circuits Technical Committee and is an editor of a number of IEEE publications. His research areas in the ECE department with his students include power and energy grids, power electronic systems, electrical machines and drives, and advanced modeling and simulation. He is part of the UBC Electric Power and Energy Systems group, whose work is focused on developing models and software for the analysis of electromagnetic transients in power systems and power electronic circuits.

To learn more about Dr. Juri Jatskevich, please visit his faculty page and researcher profile.

ECE PhD Student Wins IEEE ICC 2022 Best Paper Award

Zihuan Wang, a PhD student in UBC Electrical and Computer Engineering, received her B.S. and M.S. degrees from the Dalian University of Technology (DLUT), Dalian, China, in 2017 and 2020, respectively. Zihuan has several awards to her name, including the National Scholarship of DLUT, a UBC Four Year Fellowship, a Mitacs Accelerate Fellowship, and a Graduate Support Initiative (GSI) Award from the ECE Department. Most recently though, Zihuan won the Best Paper Award at the 2022 IEEE International Conference on Communications!

We are excited to share an interview with Zihuan below, where we connected with her about the award, her research focusing on machine learning applications in wireless systems and integrated sensing, and her future plans!

Congratulations on winning the IEEE ICC 2022 Best Paper Award! How do you feel about receiving this award?

It is wonderful to receive this award! I feel very encouraged by this acknowledgement and recognition of my work. It motivates me and strengthens my confidence in continuing my research in this area. And I am greatly thankful to my supervisor Prof. Vincent Wong who has suggested this research topic to me and provides me continual support and helpful guidance.

What topic are you researching for your PhD and what is it that excites you about this topic?

My research areas include the applications of machine learning techniques in wireless networks. In particular, I developed a deep neural network for traffic prediction in wireless networks and I tested its performance using a real-world collected dataset. I felt very excited when I first saw the greater performance improvement of my proposed approach than the other schemes, and realized how my research work can help the real-world wireless systems.

What has your career path looked like so far? How did you end up at ECE, and why did you choose our department?

I obtained my Bachelor’s and Master’s degrees in China. After that, I applied for the Ph.D study in UBC ECE. I would like to become a researcher in wireless communications in the future. To gain a wider and deeper vision in research, UBC ECE is a perfect choice for me because of its advanced resources. Also, I was interested in Prof. Vincent Wong’s research areas, and looking forward to joining his research lab.

Where do you hope you’ll be in 5 years?

I hope at that time I am still passionate about researching wireless communications. I hope to become a researcher who can contribute to the development of wireless networks.

What do you like to do outside of work/research and what makes these activities enjoyable for you?

I really love the beautiful environment in Vancouver. Outside of work, I usually go hiking. I feel relaxed during hiking, and it is a lot of fun to explore new places in nature when going on a hike. I also love watching movies, as I can see various worlds, feel different emotions, and learn new things.

For any inquiries, Zihuan can be contacted at zihuanwang@ece.ubc.ca

Meet ECE’s 2022 3-Minute Thesis Winners!

We are excited to introduce you to the winners of ECE’s departmental heat last term, and share a snapshot into their research and their experiences so far in ECE! Congratulations to Ivelina Daiss, Hooman Vaseli, and Jonas Welsch on their successful presentations!

3MT (3-Minute Thesis), is a competition held throughout UBC where graduate students are challenged to present their research in just three minutes to a non-specialist audience! Individual departments hold their own heats to determine which students from the department will go on to compete in the UBC-wide competition. Ivelina, Hooman, and Jonas were our ECE winners this year.

We connected with Ivelina and Hooman about their experiences in ECE and in the 3MT competition, where they shared their favorite parts about their programs as well as some tips for their younger selves!

What is the topic of your research?

Ivelina: I am part of the Earthquake Early Warning for the Smart City using 5G (EEW-SC) project. The project belongs to the I2Sim group headed by Dr. Jose Marti in the ECE Department and Dr. Carlos Ventura in the Department of Civil Engineering. We are designing a modern earthquake early warning application and working with Rogers Communications to verify its deployment for Vancouver. My thesis is related to providing an IoT solution for the EEW-SC application. I worked on setting up and testing a network of sensors for collecting seismic data during an earthquake.

Hooman: Broadly, I am researching machine learning and deep learning methods for medical imaging applications. Specifically, in my lab, RCL (Robotics and Control Lab), our group focuses on ultrasound images of the heart (called echo) and prostate to automate the analysis of such data and assist the clinicians. My target is to improve the stroke risk assessment through the use of deep-learning-based analysis of heart ultrasound images. So, I’d like to say I’m specializing in broken hearts! 😉

What inspired you to study this topic?

Ivelina: I feel that my work on this project is a great way to contribute to the well-being of the society I live in. In my undergraduate studies at UBC, I was introduced to Dr. Marti’s extensive expertise in systems and disaster response and management and I really wanted to join his team. In addition, working with industry partners, in particular with Rogers Communications, provided me with an experience few graduate students are able to have.

Hooman: I was first inspired to pursue a biomedical specialization during my undergraduate degree at UBC because, from my childhood, I remember how in the Star Wars universe the technology was so advanced that prosthetic arms and legs could let the user feel and control them as if they were real limbs. I got into the field of machine learning for medical imaging thanks to my current supervisor (Prof. Purang Abolmaesumi) who introduced me to the field when he offered me my first co-op job at RCL after finishing my 2nd year of undergrad at UBC. Working in such a cutting-edge field using real data to have an impact was so satisfying that I decided to continue my graduate studies as a Ph.D. student in the same lab. In 2020 when I started my Ph.D., I got to know about all the ongoing projects in RCL, and this specific project to improve stroke risk assessment by analyzing echo data seemed to me as one of the most impactful projects, especially because stroke is the 2nd leading cause of death worldwide. And honestly, this has been one of the most persistent reasons that have been driving me to do research in the field.

If you could go back in time and meet yourself at the start of your degree, what advice would you give yourself?

Ivelina: I would advise myself to take as many courses as I can. I advise future graduate students to benefit as much as possible from the extensive course offering in ECE. There are many graduate level courses with instructors who are on the cutting edge of research. Taking advantage of the knowledge they communicate is one of the best ways to gain the most from your graduate experience.

Hooman: Take it easy Hooman! Focus more on the tasks at hand rather than thinking of all the things you can do over several years at the same time so you don’t get overwhelmed. Also, celebrate every achievement you get, no matter how small. You’ll get refueled and morale boosts that you’ll need in the difficult path of research.

What made you want to participate in the 3MT competition? What was your favourite part of the experience?

Ivelina: My initial motivation to participate was to spread awareness of the Earthquake Early Warning for the Smart City using 5G project. I also felt that the preparation for the competition would help me to set in perspective what is important to communicate in my thesis. My favourite part was the preparation for the competition. I found that the Department of Graduate and Postdoctoral Studies offers extensive resources for our 3MT preparation, and I benefited a lot from the coaching sessions they offered.

Hooman: Other than the prize money, it felt really good to get a platform and present my work to people whom I had not known previously and get acknowledgment for all my efforts. Also, presenting to such an audience was well outside of my comfort zone which by itself is very rewarding. My favorite part was getting to know about all the other cool projects and their researchers in the community and making new friends.

What are you hoping to work on in the future/after your program is over?

Ivelina: I would like to apply my skills and knowledge in a way that benefits society. Working on projects in the IoT space for Smart Cities and Smart Energy and supporting new generation communications networks aligns with my perspective that our quality of life can be improved through conscientious application of technology.

Hooman: I am aiming to work in the industry after graduation, but I really hope what I’d be doing is impactful for the greater good of society rather than just making more profit for the companies. Because for me, nothing can be more satisfying than knowing what I work on makes a real difference in the world.

What non-academic experience has made your time at UBC memorable so far?

Ivelina: The people I have met, both during my undergraduate and graduate time at UBC, have made the greatest impact in my life. I am grateful for the relationships I have built with my undergraduate friends and my current colleagues. I continue to draw inspiration from them and I am so proud to see them succeed.

Hooman: Honestly, the pandemic and remote work made it really hard to get the full experience of UBC. Thankfully, I was here during my undergrad and campus has started reopening. Based on my undergrad experience and what I plan to do in the future, I’d say all the events that I’ve attended and will attend, especially the UBC signature events like Storm the Wall and the Day of the Longboat, are what make my UBC experience memorable.

Congratulations again for all your hard work Ivelina, Hooman, and Jonas!

To anyone interested in taking part in future 3MT heats, stay tuned for our call for participants early next year for the 2023 3MT Competition!

Ivelina is a student in ECE’s MASc program. To connect with her about her 3MT experience, her research, or her MASc program, Ivelina is happy to connect via LinkedIn: www.linkedin.com/in/ivelinadaiss

Hooman is a student in ECE’s PhD program. To connect with him about his 3MT experience, his research, or his PhD program, Hooman is happy to connect via LinkedIn: https://www.linkedin.com/in/hoomanvaseli

Alumni Perspectives – Sebastian Zhou, SoC Global Routing Techniques

Sebastian Zhou is a PhD alumni from our UBC Electrical and Computer Engineering department and part of UBC’s System-on-a-Chip Research Lab.

Sebastian’s PhD research focused on the optimization of System-on-a-Chip global routing techniques, and he has worked on projects developing routability-driven routing strategies. Prior to becoming a doctoral student, Sebastian worked as a Research and Development Engineer at Blackcomb Design Automation, where he designed and developed algorithms to improve hardware repairability. He has also worked as a Research and Development Technical Intern at Synopsys Inc. Sebastian was part of the PrimeTime team, helping to develop one of the best timing signoff solutions in the industry.

Sebastian joins us in an interview where he shares how he became interested in System-on-a-Chip and what excites him about his work, suggestions he would give to his younger self, as well as his talents for go-karting!

Your work mainly focuses on System on a Chip (SoC) global routing techniques. How did you get started studying this topic?

It was the start of the second year of my PhD program- I had just finished all my coursework and felt every bit of myself ready to dive into the world of research. I started working part-time at a local Electronic Design Automation (EDA) company in Vancouver. During this period, I realized that there are so many real-world problems waiting to be optimized, and there is so much room for improvement in SoC global routing. Therefore, it was at that moment I decided to dedicate my Ph.D. life to it.

Why is it important to study SoC? What future impacts could this research have?

Studying how to design better SoCs is very crucial for improving our everyday lives. Let’s think about the things that we can’t live without nowadays like cellphones and computers. They all have chips in them. Even the microwave has chips. I would argue that SoC is the foundation that supports modern human civilization. My research finds ways to accelerate the SoC production while ensuring chips are as powerful as they can be, so that (hopefully) people may get their hands on new iPhones even more frequently.

What aspects of this research did you find most exciting?

During my research, I have developed an encoding/decoding algorithm that bridges the gap between academic and industrial data formats. This algorithm enables me to work on the real chips. So, when I first saw the positive results from my experiments that my work brings to the “real world products,” I felt excited, as if I had changed the world. Of course, this conceit didn’t last long as I recovered my rationality after a good night’s sleep.

If you could go back in time and meet yourself at the start of your Ph.D., what is some advice you would give your past self?

Be more prepared, both mentally and physically. Doing a Ph.D. is nothing like doing a master’s or bachelor’s program. You will encounter hardships that no one has experienced, which means you have no one to talk to when you are down. You will face failures that you have to spend night after night solving, which means you don’t get to sleep. And even if you are in bed, you can’t sleep at all – so just be prepared. However, there are many online mental health resources that helped me throughout my degree.

You also work as a Professional Development TA Facilitator for the Faculty of Applied Science. Why is student learning important to you?

Having a good student learning atmosphere is definitely essential. I’ve been there since I was a first-time TA. I didn’t know what a TA should do, or how I should do it. I felt helpless. Although it didn’t take long to get on the right track, the experience wasn’t actually pleasant for both the students and me. I don’t want the same thing to happen to others. Therefore as a senior, I became a TA facilitator to get new TAs ready by sharing my own experience and real teaching scenarios.

Getting a Ph.D. can get pretty intense, what are your favourite things to do to unwind?

I find myself adventurous when it comes to off-work life. To get away from intensive work, I love to do intensive sports. I drive the go-kart regularly, and I am relatively competitive in this sport. By ‘relatively’ I mean I am usually the best in every day drop-in racings, and I get into the top 10 weekly rankings. However, when it comes to the prized competition and things are getting very serious with my studies, I become just a participant rather than a competitor. Now that I have graduated, maybe it’s time for me to get that trophy!

Where do you hope you’ll be in 5 years?

In 5 years, I hope that I am still working in this field, either in academia or industry. I want to proudly say to the world that some product has my work embedded in it, and that it has been benefiting the world in many aspects.

Congratulations, 2022 Graduates!

Dear Class of 2022,

Congratulations on completing your program and being a member of the ECE graduating class of 2022! This is a wonderful accomplishment, and we are proud to be your staff and faculty that have worked with you throughout your degrees. It means a lot to us that you have achieved this milestone.

To celebrate this accomplishment, we invite you to attend our graduation brunch on June 1st at 11:00 in the Kaiser Atrium. At the event we will be excited to see you and provide you with an ECE gift to remember your program by. We invite you to stay in touch with us wherever life takes you after graduation, and we are always happy to welcome you back to meet other students, attend our social events, or simply to say hello.

Again, congratulations, and all the best in the next stage after graduation!

– Steve Wilton

On behalf of your ECE Department

Sign your name and send a message to your graduating class on the 2022 virtual yearbook!

ECE Graduation 2022 Yearbook

How to Stay Connected with the ECE Community

Explore ways to stay in touch with the ECE and APSC community as an alumnus.

Complete the Exit Survey

Help us shape the student experience in ECE in the years to come! We encourage all graduating ECE students to complete the following exit survey to share a little bit about your programs with us. We hope to use these answers to continue to improve the experience of our students going forward, and your responses will be greatly appreciated!

Connect with ECE on Social Media

Join Us or Events

Or get involved with ECE Events here

Loading…

Connect with Applied Science

Connect with APSC alumni

Build your network and get started on making those important connections. A lifetime of support and opportunity awaits you!

Explore

Get involved with APSC

If you’ve got time and talent to spare, just a few hours a month can make all the difference to a student or fellow alumni.

Explore

Explore APSC career resources

Whether you’re just starting out, navigating change or completely redefining your career, APSC is committed to providing you with programs and resources to help you achieve your professional goals.

Explore

UBC alumni benefits

Explore career resources, network connections, mentoring opportunities, and discounts on travel, events, products and attractions.

Explore

Explore More Ways to Get Involved as an ECE Alumni

Explore

ECE PhD Student Mohammad Jafari, Governor General’s Gold Medal Winner

Congratulations to ECE PhD student Mohammad Jafari, who has been awarded the 2022 Governor General’s Gold Medal! This prestigious award goes to the graduating doctoral student who has achieved the most outstanding academic record in their year.

Mohammad’s high-achieving studies took place under supervisor Purang Abolmaesumi at the Robotics and Control Laboratory, where he studied deep learning for medical imaging. After graduation, he will continue his investigations into artificial intelligence as technical director of AI at Aspect Biosystems.

We spoke to him to learn more about this award, his plans for the future, and his insights and advice from his experience at ECE.

How do you feel about this award?

Indeed, it was a surprise! UBC has many outstanding students, and I feel honoured and humbled to be a recipient of this award. Special thanks goes to my PhD supervisor, Prof. Purang Abolmasumi, for his pivotal role, support, and guidance. I am grateful to the selection committee and am eternally thankful to UBC for being home to my study dreams.

What was your PhD research on, and how did you initially get started studying this topic?

My PhD research was on developing robust deep learning methods to promote reliability of computerized echocardiography. My choice of topic was foremost initiated and directed by my PhD supervisor, Prof. Purang Abolmaesumi. My topic further evolved by advancing my research and via fruitful collaborations and discussions within our research group – RCL at the ECE department, and with our clinical collaborators at Vancouver Coastal Health.

What element- experience, event, course, mentor, conference, etc- of your time as a PhD student stood out to you the most, or was most important to you?

One of the most impactful activities for me was attending academic conferences. I was fortunate that my research was recognized by the community through best paper awards and nominations, which reinforced my research motivations and affirmed to me that I was on the correct path. Furthermore, the conference gatherings helped a lot in forming external connections and maintaining a better understanding of research trends in the bigger picture.

What are your plans after graduation?

I plan to continue my research in artificial intelligence (AI) and healthcare. Currently, I work as technical director of AI at Aspect Biosystems, a Canadian company that’s pioneering 3D bioprinting and tissue engineering. I hope to make impactful contributions to this field, working towards the next generation of AI solutions and AI accelerated therapeutics.

If you could go back in time and meet yourself at the start of your academic journey, what is some advice you would give your past self?

One advice would be to simplify your goals and focus on a few high-level purposes. We can do anything, but we cannot do everything.

As well, always appreciate the importance of varied connections; a real asset of the academic journey is the chance to get connected with talented, like-minded people. Be reachable, supportive, and understanding of others.

Connect with Mohammad on Linkedin.

Making Deep Neural Networks Reliable

The wide-ranging possibilities of machine learning are transforming many industries, shaping the development of many essential applications, from self-driving cars, to healthcare, to fraud detection. But what happens when machine learning breaks down?

Like any software, deep learning applications depend on the hardware they run on – hardware that can break down and cause applications to make mistakes. Two ECE graduate students, Zitao Chen and Ali Asgari, are working to prevent these accidents. Their new software solution, Ranger, checks for and corrects faulty values in deep learning systems.

“Basically,” says Zitao Chen, the lead investigator on this project, “The problem we are looking at here is how to make machine learning models be even more reliable.”

Most computer systems are composed of software and hardware working together. However, hardware can be subject to faults such as cosmic rays, stress on the system, age, faulty designs, purposeful attacks, and so-on; all can cause a system to malfunction. When the hardware isn’t working properly, the software will be affected, and computation errors are the result. Usually, when a fault arises in a software’s computation, someone can go in and recompute it manually. But this can be difficult and laborious, and is hard to complete under time constraints, like when a self-driving car is operating.

What Zitao and Ali’s tool, called Ranger, does is anticipate and correct errors in the hardware. It performs range checks (hence the name), checking that the values of the software fall within an accepted statistical range. If there is a value that falls outside this range, it brings this erroneous value back to a safe region.

In the above example, Ranger is correcting an autonomous vehicle’s faulty computer vision. On the far left, the car recognizes the road correctly. In the center, a fault has occurred, affecting the car’s ability to infer which way to go- and directing it into traffic. In the right image, Ranger has corrected this error and the car can proceed safely.

Zitao and co-researcher Ali Asgari began developing this research at the Dependable Systems Lab, led by Karthik Pattabiraiman. “We had been working on another paper, trying to understand why the machine learning model would fail from a hardware fault,” says Zitao. “Long story short….. We came up with a nice way to characterize the patterns of when the model fails and when the model would not fail.” By identifying these patterns, they were able to develop a software that anticipated and corrected errors, bringing values back to a tolerable region.

“The very cool thing about this is that this tool is application-oblivious; so it doesn’t matter if you’re using a self-driving car or you’re using an autonomous robot- you can use this application.” says Ali.

This tool will soon be used in industry. Intel has added this software to OpenVINO – a toolkit that optimizes and improves the reliability of computer vision hardware and software. Ranger will soon be used as part of this toolkit to help develop all different kinds of deep learning technology.

“We can easily incorporate it without a lot of programming interventions.” says Zitao. “You don’t really have to make a lot of changes.” As well, “it’s really low cost…. And super effective. I think this is what makes this technique so appealing for them.”

“Despite its simplicity, it provides a very high level of reliability.” adds Ali.

This was the first project Zitao worked on after joining Karthik’s lab. Says Zitao, “I think to me the most exciting part [of this research] was understanding the problem we’re dealing with… I was playing with tools to see how to model, see how it would fail, and then, through this process, by looking at all these crazy details…gradually this emerged.”

“Once we’d identified what this problem was, we moved on to see how we can make things better based on our understanding.” he says.

“I found this research project interesting because I found it more close to the application.” Ali explains. “This technique is used in the industry and has impact. And the specific area we’re focusing on, artificial intelligence and machine learning, is being used in many applications and is ever-growing in its popularity.”

“I can see an important shift happening- formerly, maybe [developers] didn’t care much about reliability of AI applications.” says Ali. However, he sees this tool as part of a change in the industry. “I think this is a first step that could also lead to looking at reliability [in software] from many different perspectives.”

Explore the research behind Ranger here and here

Learn more about OpenVINO

Connect with Zitao and Ali

Now hiring tenure-track positions!

ECE is hiring for Associate Professor or Professor (Tenure), Canada Excellence Research Chair in Neuroprosthetics!

Sudip Shekhar awarded Killam Teaching Prize

Dr. Sudip Shekhar, Associate Professor at Electrical and Computer Engineering, has been awarded the Killam Teaching Prize.

This award recognizes excellence in teaching, and is awarded through nomination by students, colleagues and alumni. Recipients are noted for their exceptional leadership, mentorship and engagement and for the positive impact they have made on their students’ lives. “[It’s important to me] to influence students in a positive way.” Dr. Shekhar says, “[and] to provide the big picture as well as explain concepts clearly.”

Dr. Shekhar joined the ECE department in 2013. He is a recipient of the IEEE Solid-State Circuits Society (SSCS) Predoctoral Fellowship, the Intel Foundation Ph.D. Fellowship, the Analog Devices Outstanding Student Designer Award, the Young Alumni Achiever Award by IIT Kharagpur, the IEEE Transactions on Circuit and Systems Darlington Best Paper Award and a co-recipient of IEEE Radio-Frequency IC Symposium Best Student Paper Award. He serves on the technical program committee of IEEE International Solid-State Circuits Conference (ISSCC), Custom Integrated Circuits Conference (CICC) and Optical Interconnects (OI) Conference. His research interests include circuits for high-speed interfaces, silicon photonics, radio-frequency transceivers and sensor interfaces.

Of the award, Dr. Shekhar says, “I am happy and humbled [to receive this prize]. It will serve as an encouragement and a reminder to keep doing better as a teacher.”