Summer 2020
Summer 2020 www.ece.cmu.edu
U.S. NEWS & WORLD REPORT RANKINGS
FEATURE 6 Power-Full Sound Waves
UNDERGRADUATE (2020) ELECTRICAL #8 COMPUTER #2
02
06
17
CAMPUS NEWS
RESEARCH
ACADEMICS
2
Welcome from the Department Head
6
Power-Full Sound Waves
17 DeltaFS Award
4
William Sanders Appointed Dean
8
To Predict an Epidemic, Evolution Can’t
18 Air Force Partnership to Fuse AI and
Be Ignored 10 Secure and Energy-Efficient
Materials Research 21 Exploring Integrated Systems
Communication 12 Using AI to Recycle Bottles 14 Recognizing Speech Recognition
GRADUATE (2021) ELECTRICAL #9 COMPUTER #4
24
30
STUDENTS
ALUMNI
EDITOR
WRITERS
Krista Burns
Madison Brewer
22 Preventing Cyberattacks on Vehicles
28 Welcome from the Associate Director of
DESIGNER
Sarah Connor
Debra Vieira
Miranda Liu
24 Success On and Off the Court 26 Microsoft Research Ph.D. Fellowship
Alumni Relations and Annual Giving 30 Congratulations Class of 2020
Ellis Robinson Daniel Tkacik Marika Yang
Carnegie Mellon University does not discriminate in admission, employment, or administration of its programs or activities on the basis of race, color, national origin, sex, handicap or disability, age, sexual orientation, gender identity, religion, creed, ancestry, belief, veteran status or genetic information. Furthermore, Carnegie Mellon University does not discriminate and is required not to discriminate in violation of federal, state, or local laws or executive orders. Inquiries concerning the application of and compliance with this statement should be directed to the vice president for campus affairs, Carnegie Mellon University, 5000 Forbes Avenue, Pittsburgh, PA 15213, telephone 412-268-2056. Carnegie Mellon University publishes an annual campus security and fire safety report describing the university’s security, alcohol and drug, sexual assault, and fire safety policies and containing statistics about the number and type of crimes committed on the campus and the number and cause of fires in campus residence facilities during the preceding three years. You can obtain a copy by contacting the Carnegie Mellon Police Department at 412-268-2323. The annual security and fire safety report is also available online at www.cmu.edu/police/annual reports/.
Produced by Carnegie Mellon University’s Department of Electrical and Computer Engineering, June, 2020.
1
Welcome Message from the Department Head Dear Alumni and Friends, I write this as our campus and global communities continue to experience unprecedented change and challenges during the coronavirus pandemic. Over the past few weeks, I have been inspired by the efforts made by so many to maintain a sort of normalcy for our students, faculty, and staff. With our teaching initiatives now completely remote, faculty and staff have gone above and beyond their calling to support our students. From virtual game nights to Zoom coffee chats, our community has made every effort to connect during these uncertain times. C AM P U S N E W S
Throughout this magazine, you will notice continued innovative research made by our faculty and students, along with new virtual initiatives that will evolve throughout the foreseeable future. As we all adjust to a new normal, I encourage you to join me in remaining optimistic and inspired. I hope to see you
Connect with us on social media by using #CMUECE
on campus again soon.
Stay safe,
Larry Pileggi Tanoto Professor and Department Head Electrical and Computer Engineering
THE CI RCUIT
3
3
WILLIAM SANDERS APPOINTED DEAN William H. Sanders has been appointed as the next
partnerships. He’s also a well-respected collaborator
dean of the College of Engineering at Carnegie
in higher education,” said Carnegie Mellon Provost
Mellon University, effective January 1, 2020.
James H. Garrett, Jr. “I am fully confident that Bill
Mellon’s Department of Electrical and Computer Engineering in the College of Engineering, as well as a courtesy appointment in the Computer Science Department in the School of Computer Science. Sanders has spent the past 25 years of his career as a tenured professor at Illinois in the Department of
Follow Dean Sanders on Twitter at @SandersCMU
Electrical and Computer Engineering and Department of Computer Science. His research interests include secure and dependable computing and security, as well as resiliency metrics and evaluation, with a focus on critical infrastructures. He has published more than 270 technical papers in those areas. Sanders has also directed work at the forefront of national efforts to make the U.S. power grid smart and resilient. “Bill is a proven leader in engineering research
will effectively shape the vision of our College of Engineering to ensure that its unparalleled advances
C AM P U S N E W S
Sanders also holds an appointment in Carnegie
in academic and research areas will continue to rise to the challenges of our fast-changing world.” Sanders earned his bachelor’s degree in computer engineering; master’s degree in computer, information, and control engineering; and doctoral degree in computer science and engineering, all from the University of Michigan. He is an elected fellow of the Institute of Electrical and Electronics Engineers (IEEE), the Association for Computing Machinery; and the American Association for the Advancement of Science. His awards include the 2016 IEEE Technical Field Award, Innovation in Societal Infrastructure, for “assessmentdriven design of trustworthy cyber infrastructures for societal-scale systems.” Sanders is also an entrepreneur and the co-founder of Network Perception Inc.
and academia who builds strategic public-private
4
THE CI RCUIT
5
POWER-FULL Sound Waves Trillions of sensors are in our future, and they will need
powering devices with ultrasound. They launch sound
energy. Batteries are routinely used to power tiny
waves that transfer over relatively long distances and
devices, but there are other options. Piezoelectricity,
are captured by tiny piezoelectric devices co-located with
the technology that converts mechanical energy
sensors, and hence, remotely powering the sensors.
scavenge energy from movement or vibrations.
“So you have a power source somewhere, and you have all the sensors. Whenever you need to
For this reason, Carnegie Mellon researchers are
power them or interrogate them, you just send
exploring the use of piezoelectricity for smart city
this blast of sound waves to them. They receive
applications. Smart cities will rely on massive sensor networks, and the sensors in these systems will need energy. Continually replacing sensor batteries would be extremely time consuming
Because these sound waves are a bit more than 40 kilohertz—right above the audible range—they do not bother humans or animals. They can efficiently transmit over 10-30 meters, which is around 30-100 feet. Although Piazza’s system doesn’t generate a lot of
would be difficult to dispose of.
electricity, it’s enough to power small radio devices that
if you could just live off of
Gianluca Piazza
it, and they turn on,” explained Piazza.
and produce waste materials that
“It would be a lot more efficient scavenged energy. You eliminate batteries and their problems, and instead you harvest energy,” said Gianluca Piazza, professor of electrical and computer engineering and the director of the John and Claire Bertucci Nanotechnology Laboratory. There are different ways of harvesting energy. While
RE S E ARC H
into electricity, is gaining attention because it can
send and receive signals. Currently, the power source that launches the sound waves needs to be plugged in. Piazza’s team would like to further develop the system so they can launch sound waves without the need of plugged-in units. To this end, researchers at Carnegie Mellon and elsewhere are exploring novel piezoelectric materials that can be used to harvest energy, which could be beneficial for indoor communications, smart infrastructure, and implantable or wearable devices.
other researchers are extracting energy from solar, heat, and mechanical vibrations, Piazza’s team focuses on
6
THE CI RCUIT
7
TO PREDICT AN EPIDEMIC, EVOLUTION CAN’T BE IGNORED When scientists try to predict the spread of something
thousands of computer-simulated epidemics in real-
across populations – anything from a coronavirus to
world networks, such as Twitter for the spread of
misinformation – they use complex mathematical
information or a hospital for the spread of disease.
“If you don’t consider the potential changes over time, you will be wrong in predicting the number of people that will get sick or the number of people who are exposed to a piece of information.”
In the context of spreading of infectious disease,
Osman Yağan, associate research professor
the team ran thousands of simulations using
of electrical and computer engineering
steps in which the subject spreads, and use that rate to project how far and wide the spread will go. But what happens if a pathogen mutates, or information
network among students, teachers, and staff at
becomes modified, changing the speed at which it
a US high school, and a contact network among
spreads? In a new study appearing in Proceedings of
staff and patients in a hospital in Lyon, France.
the National Academy of Sciences (PNAS), a team of Carnegie Mellon University researchers show for the first time how important these considerations are. “These evolutionary changes have a huge impact,” says CyLab faculty member Osman Yağan, an associate research professor in Electrical and Computer Engineering and corresponding author of the study. “If you don’t consider the potential changes over time, you will be wrong in predicting the number of people that will get sick or the number of people who are exposed to a piece of information.” Most people are familiar with epidemics of disease, but information itself – nowadays traveling at lightning speeds over social media – can experience its own kind of epidemic and “go viral.” Whether a piece of information goes viral or not can depend on how the original message is tweaked. “Some pieces of misinformation are intentional, but
These simulations served as a test bed: the theory that matches what is observed in the simulations would prove to be the more accurate one. “We showed that our theory works over real-world networks,” says the study’s first author, Rashad Eletreby, who was an electrical and computer engineering Ph.D. student when he wrote the paper. “Traditional models that don’t consider evolutionary adaptations fail at predicting the probability of the emergence of an epidemic.” While the study isn’t a silver bullet for predicting the spread of today’s coronavirus or the spread of fake news in today’s volatile political environment with 100% accuracy – one would need real-time data tracking the evolution of the pathogen or information to do that – the authors say it’s a big step. “We’re one step closer to reality,” says Eletreby.
some may develop organically when many people
Other authors on the
sequentially make small changes like a game of
study included ECE Ph.D.
‘telephone,’” says Yağan. “A seemingly boring piece
student Yong Zhuang,
of information can evolve into a viral Tweet, and we
Institute for Software
need to be able to predict how these things spread.”
Research professor Kathleen
In their study, the researchers developed a mathematical theory that takes these evolutionary changes into consideration. They then tested their theory against 8
data from two real-world networks: a contact
THE CI RCUIT
RE S E ARC H
models to do so. Typically, they’ll study the first few
Carley, and Princeton Electrical Engineering professor Vincent Poor.
Osman Yağan
9
RE S E ARC H
SECURE AND ENERGY-EFFICIENT COMMUNICATION
Using massive multiple-input-multiple-output (MIMO)
exploration of useful device characteristics and circuits
technologies alleviates demand and increases the
architectures, we will be able to reduce the power and
efficiency of spectrum sharing for users, but requires
area significantly,” said Chen.
significant power consumption. To achieve dynamic spectrum access, Chen and her Energy-Efficient Circuits and Systems (EECS) Lab are working to build lightweight and miniaturized versions of base stations. Today, base
Vanessa Chen
Daily use of smartphones, laptops, and other Internet of Things devices requires
stations are large towers and consume a lot of power
strong, reliable, and safe wireless internet access. Because so many people need
to broadcast and receive information. With “mini base
access to use a multitude of devices, often in the same space, there has been a
stations” or “mini cells” that can be deployed as drones,
great demand for secured spectrum sharing.
underserved areas can access wireless internet.
To help meet this demand, Vanessa Chen, assistant professor of electrical and
In this project, Chen has two main goals: the first is
computer engineering, has been awarded a National Science Foundation (NSF)
to achieve low power consumption with a lightweight
grant as part of the SpecEES program to develop trusted wireless transceivers for
base station. The second goal is to enhance the wireless
secure and energy-efficient communication. Along with the real-time machine
communication security of these systems.
learning and encryption project supported by the NSF CAREER award, this new project is expected to deliver more secure spectrum access for continuous health and environmental monitoring.
10
THE CI RCUIT
Because everything is connected through the internet, it is vital to keep data protected and private. A base station can be jammed if someone sends a very high-powered signal, which is called a blocker. Chen is investigating new ways to suppress the interference of intentional jammers, such as recognizing each individual user immediately. This has proved to be quite challenging, because any new system or trusted receiver added to the base station consumes power. “Every time we try to add any function to this device, it will drain more power,” said Chen. “We need creative ways to develop security measures that are low power
“If we can enable security primitives and features in
and can be carried with a wireless device so that the
radio frequency (RF) and analog domain through the
malicious attacks can be directly tackled at the wireless front-end.”
11
In a beverage factory in Thailand, old, used glass
more recently building AI for robots in retail environments,
bottles move on a conveyer belt, heading toward their
began working with his lab on an algorithm for automatic
fate. They are on their way to be examined and put
bottle screening.
into specific categories based on whether or not they can be reused and recycled. Some of the bottles are perfect and pristine, some are chipped and scratched, while others fall somewhere in between.
bottle type, type of substance present on and in the bottle, along with other external factors like lighting variation,
The moving bottles meet a system equipped with
data capturing speed, and image quality,” said Sreena
four cameras that snap images of the bottles’ neck,
Nallamothu, project manager and researcher at CyLab
bottom, and two sides. By the time they reach the
Biometrics Center. “We mainly focus on utilizing the multi-
end of the conveyer belt, the system has used AI
view information about the texture and minor details of
to select a category for the bottle. This is the work
the bottle while being invariant to external factors to filter
of a collaborative project in partnership with CMKL
it into ‘Reusable,’ ‘Clean before reuse,’ and ‘Non-reusable’
University to developing an effective and efficient
categories.”
artificial intelligence (AI) system for accurate bottle screening for reuse and recycling, an important issue for industrial plants.
On the ground in ThaiBev’s factory, researchers gathered data on a wide variety of
In the U.S., states with container deposit legislation
bottles from each of the
(also known as “bottle bills”) have an average beverage
categories. The data was sent
container recycling rate of around 60 percent, while
to Savvides and his team to
non-deposit states only reach about 24 percent,
train the AI algorithm. The data
according to the Container Recycling Institute. But
is not constant; it changes as
in Thailand, there is no container deposit legislation.
ThaiBev’s needs and priorities
ThaiBev, Thailand’s largest beverage company, has set
change. With the various
a goal to achieve a recycling rate of 80 percent. To do
datasets, the team in Pittsburgh has trained the models,
this, they need to streamline the recycling process, and
and with easily available on-the-market cameras, the
this project will help increase efficiency.
system analyzes the images taken as a bottle moves along
“Bottles are one of the wastes that come from business, creating a lot of trash,” said Orathai Sangpetch, vice president of CMKL University. “But not all bottles can be reused—some are broken,
“What really excites me is taking the basic research of how we design pattern recognition and AI algorithms to look for defects in bottles.”
task, where we need to consider different components like
RE S E ARC H
USING AI TO RECYCLE BOTTLES
“Automated bottle screening for reusing is a challenging
Marios Savvides
the conveyer belt and makes a quick decision on which category to put it in. The project is still in progress and undergoing changes and revisions, but so far, the results have been positive.
something inside or on the bottle can’t be washed, and
For Savvides, this unique project as not only an opportunity
they need to be thrown away.”
to collaborate with CMKL University, but also a chance to
Sangpetch reached out to Marios Savvides, professor of electrical and computer engineering and director of the CyLab Biometrics Center. Savvides, who has a strong research record in the application of pattern recognition and AI for face and iris biometrics and
engineer a solution to a scientific problem in real time. “I’m excited about how we’re applying AI to solve a real-world practical problem that ThaiBev is facing in Thailand,” he said. “What really excites me even further is taking the basic research of how we design pattern recognition and AI algorithms to look for defects in bottles.”
Marios Savvides, professor of electrical and computer engineering
12
THE CI RCUIT
13
feature called “medium-time processing.” One of the
anyone with a smartphone capable of the feature knows
fundamental tradeoffs in signal analysis is choosing
it still has some way to go. Fortunately, Richard Stern, a
between a precise time measurement or a precise
professor of electrical and computer engineering, and his
frequency measurement. While most speech recognition
group research new ways to improve speech recognition
software works using short-time processing– sacrificing
software and have created an algorithm that introduces
frequency precision – Stern and Kim supplement the
a new feature: power-normalized cepstral coefficients
short-time processing with a longer time analysis they
(PNCC). This work was part of Stern’s graduate student
call medium-time processing. Using medium-time
Chanwoo Kim’s doctoral thesis. Kim now serves as a
processing, they can estimate the statistical characteristics
vice president at Samsung Corporation in South Korea.
of the background noise. These characteristics change
The IEEE Signal Processing Society, a worldwide organization committed to advancing this area of science, honored Stern and Kim with a 2019 Best Paper Award for their 2016 paper on PNCC. Stern said he was honored to be given this award.
slowly, so a precise time measurement isn’t needed, enabling a more precise frequency measurement. The PNCC coefficients are developed at both time scales: short-time for characterizing speech and longer-time for characterizing the environment. Word recognition accuracy is the most important
“There still is this real satisfaction from having an
assessment of speech recognition software. In most
idea and having the idea be right, having useful
cases, PNCC provides better recognition accuracy than
results come from that,” Stern said. “It’s very nice
other algorithms. Unfortunately, this comes with a slightly
to have that recognition, even for a moment.”
more costly computation. Computational cost is related to
Stern attempts to use knowledge of human auditory processing to guide his algorithm design, so PNCC is specifically structured to mimic some of our brain’s methods of interpreting words from soundwaves. Stern’s research, funded by the National Science Foundation and Samsung,
Richard Stern
how long it takes for an algorithm to run — an algorithm that is too costly will be too complex for real-world use. According to Stern, PNCC processing is not substantially more costly than other algorithms. Furthermore, he says the increased cost is well worth the improved results. “If you do everything with all the bells and whistles, it might be one and a half to two times the original cost,” Stern said. “But the overall cost of signal processing nowadays is just a tiny fraction of the overall computational load.”
provides a new algorithm for
Stern said the most important feature of his research is
the features of sound that
the interplay between engineered speech recognition
speech recognition software
technology and the science of human auditory processing.
uses to guess what words
have been spoken. These features are designed to better recognize words spoken in difficult environments such as additive noise, traffic, background music, speaker interference, and reverberation. Stern said he focused especially on reverberation and other people talking in the background because these are
RE S E ARC H
RECOGNIZING SPEECH RECOGNITION
Speech-to-text translation has come a long way, but
“The performance that we’re getting is a consequence of applying principles based on what we know about auditory perception to this very engineering problem of speech recognition,” Stern said. “I think that this kind of success speaks to the benefit of...bringing together information, knowledge, and wisdom from multiple disciplines.”
especially difficult conditions for speech recognition. “This is what robustness is intended to refer to,” Stern said. “Speech recognition systems should just work under any circumstance, regardless of what the acoustical environment might be.” Stern and Kim’s algorithm contains a unique 14
THE CI RCUIT
15
• 2019 •
2019 R&D 100 AWARDS IT/Electrical Winner Los Alamos National Laboratory DeltaFS—Rapidly Searching Big Data to Accelerate Scientific Discovery Brad Settlemyer, Scientist Los Alamos National Laboratory George Amvrosiadis, Research Professor Carnegie Mellon University
• 2019 •
2019 R&D 100 AWARDS Gary Grider, HPC Division Director Los Alamos National Laboratory
IT/Electrical Winner
Qing Zheng, Research Assistant Carnegie Mellon University
Los Alamos National Laboratory DeltaFS—Rapidly Searching Big Data Charles Cranor, Systems Scientist to Accelerate Scientific Discovery Greg Ganger, Jatras Professor Carnegie Mellon University
FINALIST
Carnegie Mellon University
Garth Gibson, Professor Carnegie Mellon University Brad Settlemyer,
Scientist Los Alamos National Laboratory
project recently received the prestigious R&D 100 Award from R&D World magazine, a leading resource for research scientists,
ENGINEERING PARTNERSHIPS THAT LEAD TO
engineers, and technical staff members at laboratories around the world. The annual
FINALIST R&D 100 Awards are given in recognition of exceptional
In Carnegie Mellon University’s College of Engineering, faculty and students come together from across disciplines — along with partners from industry and government — to tackle complex societal problems.
AC AD E M IC S
The Parallel Data Lab’s DeltaFS
George Amvrosiadis, Research Professor Carnegie Mellon University Gary Grider, HPC Division Director Los Alamos National Laboratory Qing Zheng, Research Assistant Carnegie Mellon University Greg Ganger, Jatras Professor Carnegie Mellon University Charles Cranor, Systems Scientist Carnegie Mellon University Garth Gibson, Professor Carnegie Mellon University
new products or processes that were developed and introduced into the marketplace during the previous year.
Explore the possibilites at makepossible.cmu.edu/engineering 16
THE CI RCUIT
17
Sitting at the nexus of data science, computer vision,
A diverse set of physical systems and research questions
and machine learning, artificial intelligence (AI) has the
will be covered by the thrust areas, but each will rely
promise to provide insights from large, high-dimensional
on several fundamental principles required to connect
datasets that can stay otherwise hidden from traditional
AI approaches to physical systems, such as enforcing a
data analysis approaches. However, its application in
physical basis in interpreting AI results and fusing rich
materials science has been comparatively slower than
and diverse datasets characteristic to materials science.
to apply AI to physical systems, as well as the wide
Air Force Partnership
to Fuse AI and Materials Research
variety of problems and data types encountered. In an effort to push forward the state-of-the-art in materials science research, Carnegie Mellon University (CMU) and the Air Force Research Laboratory (AFRL) are establishing a collaborative Center of Excellence. The center will leverage the strengths of the two institutions to develop nextgeneration aerospace materials, establish a pipeline of research talent with both AI and materials science expertise, and advance the materials science field by integrating AI into materials research and design.
Multiple colleges, schools, and institutes will be brought together across CMU through the Center of Excellence, including the College of Engineering, the School of Computer Science, and the Robotics Institute. Central to the goal of forging a deep connection
AC AD E M IC S
some fields due to the specialized knowledge required
between AI and materials science is developing a next generation of researchers with competencies in both areas. The Center of Excellence plans to support 12 to 16 Ph.D. students over its term, providing a unique educational experience where students will have access to AFRL personnel, facilities, and hardware. Each student will have the opportunity to work during the summer term at the Wright Patterson Air Force
The 5-year, $7.5M joint Center of Excellence, named
Base, a premier AFRL research facility located in
Data-Driven Discovery Of Optimized Multifunctional
Dayton, OH. The award will prepare a materials science
Material Systems (D3OM2S), is supported by an award
workforce fluent in AI methods and applications with
from the Air Force Office of Scientific Research (AFOSR)
a strong potential to contribute to AFRL in the future.
and the AFRL Materials and Manufacturing Directorate.
The center will focus on four major research thrust areas: characterization and prediction
multimodal
of rare events
data fusion
in materials
18
THE CI RCUIT
adaptive
adaptive materials
experimental
design in complex
design
environments.
19
Exploring Integrated Systems Integrated circuit design and engineering is more critical than ever. With a slowing of Moore’s Law and the end of Dennard Scaling, challenges abound to deliver ever increasing performance and power efficiency. In an effort to respond to this issue, Carnegie Mellon’s Department of Electrical and Computer Engineering has partnered with Apple to create unique opportunities for students to explore this area. Through the Apple Master’s Scholarship and Ph.D. Fellowship in
AC AD E M IC S
“We are excited to partner with Apple to offer our students this amazing opportunity.”
Integrated Systems, select students interested in IC engineering, SoC design, verification, test, and Computer Architecture have the opportunity to partner with Apple’s Hardware
Larry Pileggi, department head of electrical and computer engineering
Technology (HWT) team to develop real-world solutions. The Apple Master’s Scholarship in Integrated Systems program provides both a summer internship at Apple and financial support covering tuition costs of the yearlong M.S. degree at Carnegie Mellon University as part of the Integrated Master’s/Bachelor’s program.
The world’s greatest problems are made of many small parts.
The Apple Ph.D. Fellowship in Integrated Systems
At Carnegie Mellon we have a unique atmosphere deeply
program provides both a summer internship at Apple
native to our culture, and practice a systems approach to
and financial support covering costs of two years of Ph.D.
problem solving.
(tuition and stipend) at Carnegie Mellon University. “We are excited to partner with Apple to offer our students this amazing opportunity,” said Larry Pileggi, department head of electrical and computer engineering. “These scholarships
Through decades of working together, we’ve learned how
and fellowships will allow our students to collaborate with
to assemble the most diverse experts and solve complex
Apple engineers to obtain exposure to world-class technology
problems piece by piece.
The most complex problems. The most diverse experts. 20
THE CI RCUIT
that will greatly enhance their educational experience.” FALL 2018 // CARNEGIE MELLON ENGINEERING //
21
PREVENTING CYBERATTACKS ON VEHICLES In 2015, two security gurus remotely hacked a
“It shouldn’t be able to communicate with
Jeep Cherokee while it was being driven down
your engine. But it can. And the hackers took
the highway. The hackers blared music through
advantage of that.”
the engine. Luckily, the hackers were only demonstrating the ability to do so, the driver was in on it, and no one was hurt.
identify those unwanted connections between computers as vulnerabilities so that the automaker could be informed before an actual
computers to control everything from the radio
attack occurred.
ways to keep cars protected from attacks like the one demonstrated in 2015. First things first: they need a map of what the car’s computer network looks like.
The team tested their tool on five real-world vehicles, including a 2009 Toyota Prius and a 2017 Ford Focus. Thanks to their tool, they discovered that the Focus was found to be vulnerable to an engine shut-down attack, and
“That’s super confidential,” says CyLab’s Sekar
the Prius had an extra computer that had been
Kulandaivel, a Ph.D. student in electrical and
installed when the car was modified to all-
computer engineering. “That type of information
electric.
is automakers’ secret sauce.”
“Using this tool, we found this unexpected
So Kulandaivel and his colleagues went ahead and
computer that was transmitting on this network,
developed their own network-mapping tool. Put
and we had no idea,” Kulandaivel says.
simply, the tool can identify all of the computers on the vehicle’s network, what messages each computer is sending, where those messages are going, and where they could go if an attacker took control over one. The tool does all of this for a vehicle in under 30 minutes and less than $50 worth of hardware. It’s important to know that the hackers in 2015 were able to take control of the Jeep Cherokee by hacking into the car’s internet-connected infotainment system. Once they were in there, they learned that the infotainment system could communicate with many more things than just the radio, or the navigation system. “You would ideally think the infotainment unit should only be able to talk to other radio THE CI RCUIT
Kulandaivel and his colleagues would ideally
With vehicles increasingly becoming outfitted with to the steering wheel, researchers are looking for
22
The network-mapping tool developed by
components, like the display,” Kulandaivel says.
STUDENTS
the car’s speakers, blasted the A/C, and killed
One might argue that automakers should just share the network map of their vehicles so that it can be ensured vulnerabilities don’t exist. But vehicles can change over time, Kulandaivel says. Owners may have aftermarket parts installed, introducing new components to the network, and even degradation of parts can change how the network works. “Say a device is starting to fail over time,” Kulandaivel says. “If that causes the characteristics of the device to change the network map, then new vulnerabilities could emerge.” Other authors on the study included Carnegie Mellon Information Security graduate student Tushar Goyal, ECE Master’s student Arnav Kumar Agrawal, and ECE professor Vyas Sekar.
23
SUCCESS ON AND OFF THE COURT Like many young engineers, second-year ECE major Vivian Beaudoin is drawn to the versatility of technology, especially computers. She loves problemsolving and played with Legos and puzzles growing up. However, Beaudoin isn’t just an engineer — she’s
STUDENTS
also a member of CMU’s volleyball team. Beaudoin fell in love with computer science in high school after taking a class at the suggestion of her older brother. She went on to take every class offered
Beaudoin celebrates with teammates during a match against Muskingum in September 2019. Photo taken by Sports Information Staff.
and looked for a college to continue her technological education. Around the same time, at the suggestion
Beaudoin said being on the volleyball team is a lot of work,
“[The class] can be really overwhelming and really
of a coach, she tried her hand at volleyball, finding an
but since she enjoys playing so much, it’s worth it. She’s
stressful,” Beaudoin said. “I wanted...to help people in
instant match.
already learned a lot from her time on the team, but she
that introductory class.”
“It was immediately the best sport ever, in my opinion,” Beaudoin said. “I very quickly kind of focused on it.”
management.
As for the future, Beaudoin has a difficult choice ahead. She must choose between a concentration in hardware
“[Volleyball] forces you to time manage...because you have
systems — the physical makeup of the computer —
a commitment of practice one to two hours every single
and software systems — the code used to make it run.
day, and you’re traveling on the weekends,” Beaudoin said.
Regardless of her concentration, Beaudoin has one thing
“You need to plan and you need to prioritize.”
in mind: helping people.
tournament wins. So she reached out to CMU’s team
Other important skills she’s learned include teamwork and
“I always find really interesting technologies...that can
and the coaches came to see her play.
adaptability.
help people who have disabilities,” Beaudoin said. “I think
“It just fell into place from there,” Beaudoin said.
“Being able to learn very quickly from mistakes is
As a student, CMU was attractive to Beaudoin because of its renowned engineering program. As a volleyball player, Beaudoin was impressed by CMU’s
Vivian Beaudoin
said the best lesson she’s learned has been about time
Since coming to CMU, Beaudoin has succeeded on and off the court. Beaudoin earned the top grade in Structure and Design of Digital Systems, a course
that would be something cool to work with.”
something that’s really important,” Beaudoin said. “Also...
In addition to the potential to help people, Beaudoin said
volleyball is a big team sport...but if they’re not working
the endless possibilities are another aspect of technology
together, you’re not going to be as successful.”
she loves.
required of all ECE majors. In the same semester, she
In addition to being on the volleyball team, Beaudoin is
“It’s really a good tool for a lot of things,” Beaudoin said.
was an honorable mention regional All-American and
involved in the Society of Women Engineers and Women
“We’re seeing new technologies that can do so many
received 2nd team All-UAA honors for her prowess
in ECE. She is also a TA for an introductory programming
things that we never imagined.”
on the court.
class. In this fast-paced course, students learn to code, some with no prior experience. As a student, Beaudoin said the staff — professors and TAs — were an invaluable
24
THE CI RCUIT
asset to success.
25
THE BRAIN POWER BEHIND THE POWER GRID Microsoft Research Ph.D. Fellowship
Jingxian Wang, a current electrical and computer engineering Ph.D. student, has received the prestigious 2020 Microsoft Research Ph.D. Fellowship. Microsoft Research awards this two-year fellowship to Ph.D. students at North American universities pursuing research aligned to the research areas carried out by Microsoft. “I am honored to receive this fellowship,” says Wang. “This fellowship provides me the opportunity to connect with top researchers in the industry which has the potential to polish my research vision. I hope my research could have more impact on our daily life in 2030.” The Microsoft Research Fellowship includes tuition and fee coverage for two academic years, a living expense stipend, an invitation to interview for one salaried internship in 2020 with leading Microsoft researchers working on cutting-edge projects related to the recipient’s field of study, and an invitation to the Ph.D. Summit: a two-day workshop held at one of Microsoft Research’s labs where fellows will meet with Microsoft researchers and other top students to share their
Protecting the power grid – Brought to you by ECE
research. Wang’s research lies in the area of the Internet of Things (IoT) with a particular emphasis on designing low-cost, long-range and battery-free networks and services. He has designed and built a series of wireless systems that elevate today’s batteryless technologies by extending communication range, enriching sensing capabilities, and enabling novel applications, specifically in the accessibility domain.
Power. Our most precious asset. You rely on it for everything – so shouldn’t we do everything to protect it? Electrical and computer engineers have developed a simulation approach – based on integrated circuit techniques – that could potentially help experts model and simulate the power grid more reliably, thereby protecting it from possible cybersecurity threats in the future. More at ece.cmu.edu
“I want to develop wireless technologies that connect everyday objects to the Internet,” says Wang. “My research so far is all about exploring different ways to make that easier and possible and developing various
ECE – THE FUTURE IS WHAT WE DO.
services that I can build on top of that.” 26
THE CI RCUIT
27
BOB STRATTAN EE ’ 5 9 , ’ 62 Alumni Message from the Associate Director of Alumni Relations and Annual Giving
Bob has fond memories of his doctoral research at CMU: adapting an electrical power systems simulator lab.
Dear Friends, I hope this finds you well and safe during these uncertain
Like all other higher education institutions, Carnegie Mellon University transitioned to online learning quickly. Our students and faculty adapted to this new normal and the spring semester was completed on time. The resiliency of our community has been encouraging. Although the
Margaret Noel
future semesters may bring challenges with in-person lectures, I am confident that our department will continue to acclimate and thrive in this new landscape.
The Department of Electrical
I know all of you have had to adjust to a new lifestyle, and I
and Computer Engineering, in
was encouraged to hear from so many alumni about their
partnership with the School of
work from home situations.
Computer Science, is excited
Looking forward to the 2020-2021 academic year, we are
alumni volunteering platform.
a strong presence at the virtual Grace Hopper Celebration
Opportunities for alumni to
held September 29 – October 2, 2020. We are also making
engage, volunteer, and connect
plans for virtual alumni engagement. If you have a specific
with current students, alumni,
idea for these events, please let us know.
and the department will be
As our recent graduates navigate uncertain employment
listed here. From speaking at a
waters, I welcome any opportunity that you may know of to
student organization dinner to
help connect our students to employers. Please feel free to
sharing advice through resume network, it’s now easier to
department. I hope to see you back on campus once it’s
sign up for opportunities.
safe for us to be together again. Please stay in touch.
Margaret Noel
Bob and his wife Vada have created a lasting legacy at Carnegie Mellon through a charitable trust that will provide support to the Department of Electrical and Computer Engineering where it is most needed.
reviews on our ECE/SCS advisor
I remain inspired by those who continue to support the ECE
Best,
Bob is one of only a few faculty members inducted into the University of Tulsa College of Engineering and Natural Sciences’ Hall of Fame.
to introduce Involve, an
making preparations for numerous virtual events, including
reach out to me directly to initiate this discussion.
After graduating, Bob worked for Boeing and North American Aviation, and then joined the faculty of the University of Tulsa for 30 years.
AL U M N I
times.
They give back because Bob says Carnegie Mellon was the pathway to a successful and rewarding life.
Visit www.ece.cmu.edu/alumni to learn more, sign up, and help inspire our future ECE alumni. Questions? Contact: involve@andrew.cmu.edu or
G I VE STRATEG ICALLY, SU P P O RT G EN EROUSLY. Learn how easy it is to achieve your philanthropic vision through a planned gift by visiting giftplanning.cmu.edu.
As a retired engineer, Bob still enjoys spending his time tinkering. He restores vintage cars, such as his father’s 1946 Chevrolet pickup.
Contact the Office of Gift Planning today at 412.268.5346 or mkoster@andrew.cmu.edu.
margaretnoel@cmu.edu 28
THE CI RCUIT
29
Congratulations Class of 2020! On Sunday, May 17, the CMU community came together around the globe to celebrate the graduates virtually. Visit www.ece.cmu.edu/news-and-events/diploma.html to view the message to the Class of 2020.
Bill Nace, Teaching Professor
Yorie Nakahira Assistant Professor
Gianluca Piazza, Professor
Mark Craven, Undergraduate
Giulia Fanti, Assistant Professor
AL U M N I
John Cohn Ph.D, IBM Fellow , MIT-IBM Watson AI Lab
Academic Program Advisor
30
THE CI RCUIT
31
Carnegie Mellon University 5000 Forbes Avenue
Pittsburgh, PA 15213 www.ece.cmu.edu
Connect With Us! Connect with us on social media by using #CMUECE /CarnegieMellonECE /CarnegieMellonECE /school/cmu-ece @CMU_ECE /cmuece