ECE's Approach for the Fall Term

ECE’s Approach for the Fall Term

Dear ECE community,

Following guidelines from UBC and our provincial health leaders, we have decided that classes for the fall 2020 term will be delivered 100% remote for the first term, with a full suite of course offerings. The department will be offering all its courses in a remote learning format. The fall term will start as scheduled, and all classes will be offered, albeit some in a revised manner. You will be able to complete the entirety of the term from anywhere in the world. We are committed to ensuring you stay on track towards your degree despite these necessary adjustments.

Within ECE, an expert team of faculty members and staff have worked diligently since mid-March to make sure new and returning students will get a solid and high-quality teaching experience and remain engaged in both academic and non-academic settings. We recognize the importance of the hands-on experience you receive participating in ECE-related extracurriculars on campus, and we’re looking into ways of providing students with a rich experience outside of the classroom as well. We hope to soon be able to share more concrete details regarding courses, registration, and extracurriculars, but for now, we will continue to work hard to shape the upcoming semester.

Be sure to stay connected with ECE and APSC through following our website and our various social media channels. We will provide regular updates, so that you can make informed decisions for next semester. If you have any further questions or wish to get in touch directly, feel free to contact us at help@ece.ubc.ca.

Steve Wilton

Department Head, Electrical and Computer Engineering

ECE Capstone Faculty Award Recipients

UBC ECE’s 2020 spring cohort’s perseverance and persistence has paid off, and they’ve seen their Capstone projects to completion. Each team overcame many challenges to provide timely deliverables spanning a wide array of ECE-related subjects. Over 230 students formed 48 teams that leveraged four years worth of learning in order to design solutions to challenges proposed by industry and community partners.

ECE is proud to present this year’s Capstone Faculty Award Winners:

Propbot (Divya Budihal, Jack Guo, Nancy Hong, Zhaosheng Li, Hannah Sawiuk)
Project Skynet (Peter Deutsch, Arthur Hsueh, Ardell Wilson, Muchen He, Meng Wang)
Procedural Generation Tool (Ian McCall, Matthew Berends, Mitch Duffield, Mathew MacDougall, Simong Song)
Digital Health & Wellness: Video Fall Detection using Deep Learning (Mohamed Hamdan, Alessandro Narciso, Abdul Moiz, Winnie Gong, Kirsten Kwan)

Congratulations to all Capstone students on your hard work over this past year.

Propbot, sponsored by the UBC Radio Science Lab

Divya Budihal, Jack Guo, Nancy Hong, Zhaosheng Li, Hannah Sawiuk

A rendering of Propbot.

Propbot aims to provide researchers with key information to design robust and transformative communication systems—and it does so completely autonomously. This team of fourth-year students created a robot architecture capable of traversing the UBC campus and completing large-scale data collection. The restrictions of COVID-19 posed a serious obstacle to the Propbot developers; to continue testing, they constructed a completely virtual environment to simulate Propbot and ensure its functionality. The team’s development of Propbot is only the first phase in a multi-year project that will shape the future of communication systems at UBC.

Project Skynet, sponsored by the UBC ECE System-on-Chip (SoC) Lab

Peter Deutsch, Arthur Hsueh, Ardell Wilson, Muchen He, Meng Wang

A Capstone student flying their team’s project—a drone capable of machine learning on the fly.

Generic, off-the-shelf component architectures available in CPUs and GPUs are not optimized for the ever-expanding field of machine learning. This team focused on showcasing the promise of tailor-made hardware being developed by Dr. Mieszko Li and team at the UBC SoC Lab. One common method of achieving this custom performance is through the use of field-programmable gate arrays (FPGAs), a kind of reconfigurable digital circuit, allowing for rapid iteration of hardware designs.

The team mounted an FPGA and the necessary circuitry on a drone equipped with a camera and CPU. The final project is capable of performing machine learning—quite literally—on the fly. The drone performs object-detection using YOLOv2, a popular open-source machine learning algorithm. The team faced countless obstacles due to the COVID crisis, such as trouble ordering parts or having to build each subcomponent separately in respect of social distancing regulations. However, they persevered and ended up with not only a fantastic final product, but also a feeling of preparedness for the unforeseen challenges in a career of engineering.

Procedural Generation Tool, sponsored by Blackbird Interactive

Ian McCall, Matthew Berends, Mitch Duffield, Mathew MacDougall, Simong Song

A screenshot of terrain generated by the team’s tool in Godot.

While the current pandemic may have forced UBC students and staff indoors, these students were already developing new ways of creating procedurally-generated virtual environments for us to explore from home. “Procedural generation” refers to the use of algorithms to create entire landscapes using various noise patterns and simulation. This team created a tool which eases the process of designing and fine-tuning those very algorithms, allowing their client, Blackbird Interactive, to more rapidly construct the worlds they envision. This, in turn, will reduce costs and provide an effectively-infinite number of high-quality environments for the studio’s fans to explore.

The software-based solution was developed in C++ and comes equipped with an editor to allow developers to preview their results. The final result can be exported and works with a variety of engines. It even ships with 33 types of terrain generation functions, right out of the box.

Digital Health & Wellness: Video Fall Detection using Deep Learning, sponsored by TELUS

Mohamed Hamdan, Alessandro Narciso, Abdul Moiz, Winnie Gong, Kirsten Kwan

The team had to create their own dataset for fall detection, and with it trained a model with an accuracy of over 90%.

Remote health monitoring is of tremendous value to the field of healthcare, improving quality-of-life for patients and greatly reducing costs in the healthcare industry. This team looked to tackle fall detection, which is of great value for the elderly and people with disabilities. Additionally, fall detection can be a life-saving signal in the event of a heart attack or stroke. Leveraging the latest advancements in machine learning technology, the team created their own labelled data set and trained a deep learning based neural network. The final model detects falls with over 90% accuracy from the video feed.

Though COVID-19 restrictions prevented in-person meetings, the team was able to do a live-demo at their last in-person meeting, as they had completed their project ahead of schedule. Their impressive accomplishment will surely support the healthcare system in preventing lives from being lost.

All projects completed in this year’s Capstone course were exemplary and showcased the grit and character of UBC engineers, while also giving back to UBC’s local community and industry. Congratulations to all who participated, and a special congratulations to this year’s award winners.

UBC Researchers Design Ultrasound Scanner Network for COVID19

Scanner expected to speed up diagnoses in rural and remote areas

UBC researchers are collaborating with local partners to establish a network of portable, handheld ultrasound scanners that can soon accelerate COVID-19 diagnosis in B.C. and potentially beyond.

The scanners pair a locally-developed commercial ultrasound device with a secure online library of lung ultrasound images and a specially developed artificial intelligence (AI) algorithm, allowing health care practitioners to diagnose COVID-19 at the point of care—almost instantly.

Family doctors and acute care units in rural B.C. will be the first users, with 50 units ready for deployment. More than 30 additional scanners will be distributed to urban acute care sites managed by Vancouver Coastal Health.

The project is co-led by Dr. Oron Frenkel, an emergency physician at St. Paul’s Hospital and a clinical assistant professor at UBC’s faculty of medicine; Dr. Teresa Tsang, UBC cardiologist and professor of medicine and director of echocardiography at Vancouver General Hospital and UBC Hospital; Dr. Purang Abolmaesumi, professor of electrical and computer engineering; and Dr. Robert Rohling, professor of electrical and computer engineering and mechanical engineering.

“With this scanner, we can potentially detect COVID-19 lung changes earlier while waiting for lab test results,” says Tsang. “This may also enable us to anticipate who will likely deteriorate rapidly, so that we can support these patients optimally from the start.”

Data from the field suggests that the scanner can detect up to 33 per cent more cases of COVID-19 pneumonia than some current lab tests. “It’s easy to use, so even physicians with less experience can obtain fast, accurate results,” said Tsang.

The team will build Canada’s first ultrasound library for lung disease and will use AI to enable the handheld scanners to accurately detect patterns typical of COVID-19 and other lung diseases at the point of care.

“This project demonstrates UBC’s expertise in applied AI research,” said Dr. Purang Abolmaesumi, the Canada Research Chair in Biomedical Engineering at UBC. “With these scanners, we showcase UBC’s and B.C.’s cutting-edge capabilities in developing AI technology for medical imaging, with direct impact on our community and the Canadian health care system.”

Dr. Robert Rohling, who also leads the Institute for Computing, Information and Cognitive Systems at UBC, highlighted the different contributions of the project members and partners. “Providing accurate, timely diagnostics for COVID-19 is a tremendous challenge. What really helps to solve it is the diverse and talented team. Each member is a leader in their field but more important is that doctors are working with engineers and UBC is working with B.C. companies.”

The scanners—called PoCUS, for point-of-care ultrasound— were designed and provided by Burnaby-based Clarius Mobile Health. They can be disinfected easily between patients and come with a mobile phone app for ease of use.

Pivoting in the midst of COVID-19

The Clarius scanners have been in use since 2017 but were swiftly adapted in March to diagnose COVID-19 in order to contribute to the public health response to the virus.

It is part of Intelligent Network Point of Care Ultrasound (IN-PoCUS), a $2.5 million project led by B.C.’s Digital Technology Supercluster aimed at improving health care diagnosis in rural B.C.

The Digital Technology Supercluster solves some of industry’s and society’s biggest problems through Canadian-made technologies. It brings together private and public sector organizations of all sizes to address challenges facing Canada’s economic sectors including healthcare, natural resources, manufacturing and transportation.

Other funding and in-kind contributions were provided by Providence Health Care, Clarius Mobile Health, Change Healthcare, UBC, Vancouver Coast Health and Rural Coordination Centre BC.

Bo Fang Wins Award in Dependable Computing!

Bo Fang, an ECE Ph.D. student has been awarded the William C. Carter PhD Dissertation Award in Dependability for his Ph.D. thesis, titled Approaches for Building Error Resilient Applications. The William C. Carter PhD Dissertation Award in Dependability is the most prestigious award a Ph.D. student can receive in the dependable computing field. It is presented annually at the DSN Conference since 1997 to one Ph.D. student worldwide. The award recognizes an individual who has made a significant contribution to the field of dependable computing throughout their graduate dissertation research. Fang says “it is an honor to be awarded by the DSN community. I am humbled to be recognized by the community and my colleagues.”

The award is sponsored by IEEE TC on Dependable Computing and Fault Tolerance (TCFT) and IFIP Working Group 10.4 on Dependable Computing and Fault Tolerance (WG 10.4) to commemorated the late William C. Carter who was a pioneer in the formation and development of the field of dependable computing.

Fang is supervised by Dr. Karthik Pattabiraman and Dr. Matei Ripeanu. His research focuses on the effect hardware faults have on high-performance computing systems. Fang’s research proposes an error propagation model and crash model to identify which faults have the potential to cause silent data corruption and crashes, allowing for selectively triggering recovery. This stems from the idea that most transient hardware faults do not have a significant impact at the software layer. Ignoring faults that do not create problems allows HPC systems to be more efficient. Fang additionally proposes applying a roll forward recovery scheme in standard checkpoint/restart systems. This trades confidence in results for efficiency in performance and energy saving.

Bo Fangs research relates to the award as the “research focuses on designing approaches for building error-resilient applications, in the context of high-performance computing scenarios.” This is tightly in line with the research performed by the dependability community. His work has been published in top tier venues, and inspired many other researchers to write follow up papers based on his research. Bo’s work has been adopted by two national labs, The Pacific Northwestern National Labs (PNNL) and Los Alamos National Labs (LANL) as well as companies such as Nvidia and AMD. Bo is a recipient of the NSERC Post-Doctorial Fellowship and was ranked number two in the computer science division. He is currently doing a post-doc at the Pacific Northwestern Nation Labs (PNNL).

Dr. Pattabiraman states the significance of Fang’s work allows “High-Performance Computing (HPC) systems [to] be much more efficient in terms of performance and energy when it comes to providing fault-tolerance. The latter is especially important as these systems consume large amounts of energy for their operation and hence Bo’s work provides significant cost-savings in these systems”

ECE Prof Kruchten Awarded 2020 Linda M. Northrop Award

Congratulations to Philippe Kruchten, an ECE professor that has been selected as the recipient of the Linda M. Northrop Software Architecture Award!

The award is given to an individual or team that has used software architecture to significantly improve practices, outcomes, or both in an organization or in the software-development community. Kruchten will deliver a webcast, introduced by Linda Northrop, called “Software Architecture: A Mature Discipline?” at 1 p.m. EDT on Tuesday, June 2.

Kruchten has been a significant member of the international software architecture community for more than 30 years. Specializing in large, software-intensive systems design, Kruchten directed the development of the Rational Unified Process, an iterative software development process framework, and developed the 4+1 Architectural View Model. His contributions as a pioneering practitioner, a thought leader, an author, and an educator have advanced the recognition of software architecture as an important topic for practitioners.

“I’m truly honored by this award,” said Kruchten. “For the last 30 years I’ve always considered the SEI as the world beacon of software architecture. So many great developments and community initiatives originated there, and much of it under the leadership of Linda Northrop. This means a lot to me.”

Original story

MASc Student Selected to Attend 2020 Heidelberg Laureate Forum!

ECE MASc student, Aarti Kashyap has been has been recently invited to participate in the 2020 Heidelberg Laureate Forum (HLF) in Germany. Established in 2013, HLF brings forth 200 young researchers worldwide in the mathematical and computer science field. Participants are carefully selected by a distinguished panel of experts to network with top people of the discipline such as Nobel Laureates and Fields Medal winners.

As many of the attendees are either senior PhD students or junior faculty and post-docs, Aarti’s invitation is quite significant. “It is very exciting. Given the fact that I am a Masters student, I did not expect to be selected for HLF,” Aarti says.

Aarti works under the supervision of Professor Karthik Pattabiraman, specializing in building formal models for safety-critical systems such as artificial pancreas system and air traffic control management. This is important as safety-critical systems have human life depending on them and hence require hard guarantees before they can be deployed. These hard guarantees can be provided through the means of mathematics.

Building formal models help in designing safer and more robust systems. Cybercrimes are rising very quickly due to the extreme dependence on computers. To protect them we need strong reasoning mechanisms such as math, making this topic extremely useful in society.

As HLF is an event that tries to bridge the gap between mathematics and computer science, this overlaps well with Aarti’s current and future research goals. This symposium provides a space for ideas and innovation, meant to motivate and inspire the next generation of scientists. Aarti says she is most looking forward to interacting with researchers who have built the mathematical foundations for Computer Science as well as meeting peers that will also be attending the event.

UBC’s Spring Graduation Ceremony

Graduation for the Class of 2020 will be held on Wednesday, June 17, 2020. There will be one ceremony for students graduating from UBC Vancouver and one ceremony for students graduating from UBC Okanagan. The virtual ceremony will include many elements of a traditional ceremony and some unique ones as well. More details of the virtual ceremony will be shared online in the coming weeks. Please visit graduation.ubc.ca for more information. UBC is committed to holding an in-person graduation ceremony for the Class of 2020 when it is safe to do so.

Read Santa J. Ono full message

ECE Researchers Join the Fight Against COVID-19!

Researchers from the Department of Electrical and Computer Engineering at UBC have joined the fight against COVID-19. They’ve developed a systematic feedback strategy they say can help public health authorities in their efforts to contain the virus over the next several months.

Their proposed methodology — inspired by work of epidemiologists at Imperial College and others — does not need to rely on accurate predictive models. It uses hospital ICU capacity as a barometer for determining when physical distancing should be tightened up, and when it should be relaxed.

The team that performed the analysis include ECE’s Guy Dumont, Greg Stewart and Klaske van Heusden. They explain the rationale behind the method as well as the importance of this research.

What is the significance of a feedback-based strategy for fighting COVID-19?

There are signs that Canada is making progress on flattening the curve of new COVID-19 infections, but public health authorities stress that it’s critical to keep the momentum going. Canadians need to continue physical distancing — the primary non-pharmaceutical weapon in this fight — over the next several months while awaiting a vaccine.

However, people need to know what their exit strategy is. When can distancing be safely relaxed for society to continue functioning? Their methodology can enable decision makers to fine-tune the timing, duration and scope of intervention measures like isolation and quarantining.

This can help public health officials bring the outbreak under control and manage hospital caseloads as the public waits for herd immunity to take effect or for a vaccine to be developed — while at the same time permitting safe relaxing of physical distancing.

What does the feedback-based method look like?

A standard SEIR (susceptible, exposed, infected, recovered) epidemic model was used, typically used by public health researchers to predict the spread and impact of an outbreak. They use the number of available hospital ICU beds as the primary measure of health care capacity. The goal is always to bring infectivity rates to manageable levels.

As an example, if hospitals in jurisdiction X are approaching overcapacity, the feedback-designed policy will suggest an increase to the physical distancing in the region. When the healthcare capacity increases, the policy can suggest an optimal time for policymakers to relax or lift these intervention strategies. It’s key that many or most interventions have intermediate options that can be leveraged and thus avoid oscillations and repeated outbreaks.

Their approach emphasizes the important role that feedback can play to stabilize the system. Left on its own the epidemic is unstable, i.e. it grows exponentially. By applying a basic control principle known as feedback stabilization, it is possible to bring and maintain the propagation rate to a level manageable by the healthcare system. Thus, they’ve drawn on engineering principles to provide policy suggestions that take into account economic considerations and medical constraints.

How does this add to the current knowledge of the novel coronavirus?

Current epidemiological models of COVID-19 do not have an accurate way of estimating when to relax and when to tighten distancing interventions. An overly aggressive on-off approach may lead to unmanageable swings in health care capacity and the number of new cases. This is for instance what happened in St. Louis during the 1918 Spanish flu pandemic.

The team believes that by bringing computer feedback to bear on the policymaking process they can have greater health outcomes for everyone concerned. They hope to work with other researchers in Canada or elsewhere to further develop this methodology, and possibly make it more interactive to help educate the public.

Hossam Shoman Wins 1st Place in ECE Heat’s 3MT Competition!

In February, Hossam Shoman won first place in ECE Heat’s 3MT Competition and has also secured a spot in the UBC 3MT Semi-Finals for his presentation, “A Stable Laser, without an Isolator”. His research began in September 2016, which focused on improving the performance of electronic chips using photonics. Photonics is a term that combines optics (photons) and electronics. As the speed of data communication on electronic chips starts to saturate, integrating optics with electronics on the same silicon chip – making the so-called photonic chip – can push these speeds beyond what electronics can do on their own. The first element required in a photonic chip is an integrated laser. Following the laser, is an isolator, a device that provides stability to the laser’s operation. These isolators are made of rare earth elements, which are bulky, expensive and do not easily integrate on silicon platforms. This has been a major blockade for the large-scale integration of photonic chips and several applications that rely on on-chip solutions.

Currently, he and his team have found a cost-effective way to stabilize semiconductor lasers. Instead of using rare earth elements to make optical isolators, they design an optical circuit, on the same silicon platform that contains the photonic components, that can perform similarly to an isolator. The way they stabilize the laser is much more cost-effective compared to the current methods. This means that lasers used in data centres and high-speed communication systems can be produced at a large-scale and low-cost.

This is significant as their technology does not rely on any changes to the current CMOS foundries that fabricate electronic chips, which is a requisite for the large-scale deployment of electronic-photonic chips on silicon platforms, and a key enabler of several revolutionizing technologies that rely on on-chip solutions, such as future quantum communication and computing systems.

Currently, they have a proof-of-concept of their working technology. A patent for their work has also been submitted, and they are in the process of writing a manuscript to publish their findings. The next step is to do market research and eventually set up a company to commercialize their technology.

Ahmed E. Mostafa Wins 2nd Place in ECE Heat’s 3MT Competition!

Ahmed Elhamy Mostafa joined the ECE department in September 2015 and officially started his research about “Medium Access Control and Resource Allocation for Massive IoT” in January 2016. He recently won second place in the ECE Heat portion of the 3 Minute Thesis competition for his presentation surrounding his research called, “The Middle-person in the Smart City”.

Essentially, the goal of his research is to enable the Internet of Things (IoT) in 5G communication systems. This can be done by providing channel and power allocation algorithms to support the IoT applications that require providing energy-efficient connectivity to a large number of devices simultaneously. These algorithms achieve the goals of energy-efficiency and massive connectivity by combining communication technologies (e.g., non-orthogonal multiple access (NOMA) and backscatter communications) and mathematical tools (e.g., optimization and machine learning).

The key finding of his team’s research is to design channel and power allocation algorithms based on maximizing the connectivity, which means maximizing the number of IoT devices that meet a certain performance requirement (e.g., minimum data rate threshold). In these solutions, IoT devices are encouraged to share communication channels using NOMA. The research proposes an algorithm that matches the IoT devices that can share the same channel while satisfying performance and power budget requirements and taking the deployment and channel status information into account.

This is important because in human-to-human communications (H2H), main applications, such as video streaming, require a higher communication quality (i.e., higher data rate). In their research, they highlight that for some categories of IoT applications (e.g., environmental sensing), providing a higher communication quality (i.e., higher data rate) is not the major objective. It is more important to provide a minimum communication quality for the largest number of devices, and so, he and his team primarily develop channel and power allocation algorithms that are more suitable for these IoT applications accordingly.

His team’s research shall encourage the industry to adopt NOMA and backscattering in the 5G communication standards to support massive IoT. In addition, some mobile applications may be updated/developed to help with IoT data transfer to provide energy-efficient connectivity for the IoT devices.

The team’s next steps include enhancing their proposed algorithms, which are built using mathematical models, by utilizing the data from IoT networks to develop data-driven algorithms using machine learning tools (e.g., deep learning and deep reinforcement learning). Ahmed has also earned a place to compete in the UBC 3MT semi-finals taking place on March 10!