Skip to main content

New wireless charging system from UW researchers uses a laser to beam power to your cellphone

Demonstration of the laser wireless charging system with a smartphone

UW researchers’ laser-powered charging system can wirelessly charge a smartphone from across a room. Guard beams (illuminated here in red for visibility) prevent human contact with the charging beam by triggering the beam’s automatic shut-off.

Mobile devices and WiFi have given people the freedom to access information and accomplish tasks on the move, but there is still one area in which we remain tethered to the old way of doing things: the power cord. No matter what fancy features or sleek designs are built into the latest devices, they still run on a battery — which means they have to be plugged into the wall to recharge. But that could change thanks to a team of researchers led by Allen School professor Shyam Gollakota and Physics and Electrical Engineering professor Arka Majumdar, who have developed the first laser-powered wireless charging solution to safely and quickly charge mobile devices from across a room — potentially liberating us from those pesky power cords for good.

The team set out to create an end-to-end power delivery system that is practical and efficient as well as safe. To satisfy the first two criteria, the researchers aimed to adhere to the form-factor requirements of a smartphone while minimizing the need for instrumentation of the environment. The resulting system consists of two main components: a laser power source capable of delivering a charge at distance, and a compact receiver that attaches to the back of a smartphone. Power comes from a laser emitter configured produce a focused beam in the near-infrared spectrum. That beam charges the smartphone via a small array of photovoltaic cells mounted on the back of the device to convert the beam’s optical power to electrical power.

To prevent overheating, the cell array is paired with a thermoelectric generator that is mounted on an aluminum heatsink to dissipate excess heat from the beam. The generator harvests some of the heat that would otherwise be wasted. Instead of requiring the laser emitter to scan the room for devices, which is inefficient and time-consuming, the team developed an acoustic localization system that relies on a series of chirps emitted by the device, in this case a smartphone, to be charged. These chirps are inaudible to human ears, but capable of being picked up by the microphones built into the emitter.

The receiver that mounts on the back of a smartphone

The receiver, which consists of a photovoltaic cell array attached to a thermoelectric generator mounted on an aluminum heatsink, is only 8 mm thick and 40 mm wide.

“This acoustic localization system ensures that the emitter can detect when a user has set the smartphone on the charging surface, which can be an ordinary location like a table across the room,” explained Vikram Iyer, a Ph.D. student in electrical engineering who works in the Allen School’s Networks & Mobile Systems Lab. Iyer is co-lead author of the research paper presenting the team’s work, which was published in the Proceedings of the Association for Computing Machinery on Interactive, Mobile, Wearable and Ubiquitous Technologies (IMWUT).

While power delivery and heat dissipation presented interesting technical challenges, the team also had to contend with the significant question of human safety. Due to the level of concentrated power required from the laser to deliver a strong enough charge wirelessly and from distance, the team had to devise a foolproof way to prevent humans and objects from inadvertently coming into contact with the beam.

“Safety was our focus in designing this system,” Gollakota said. “We have designed, constructed and tested this laser-based charging system with a rapid-response safety mechanism, which ensures that the laser emitter will terminate the charging beam before a person comes into the path of the laser.”

The team could not rely on cameras, acoustic tracking, or other motion detection methods, because those approaches would all introduce an unacceptable delay between detection and shut-off. Instead, the team turned to optics, incorporating a set of “guard beams” — harmless, low-power lasers that surround the charging beam. A set of 3D-printed retroflectors arranged around the power cell reflect the guard beams back to photodiodes on the laser emitter. When one of the guard beams detects an object, such as a human arm, moving into its path, it causes the charging beam to automatically shut off. Because the guard beam operates at the speed of light, it can trigger this shut-off within a fraction of a second — faster than the maximum speed of human motion.

Team members, clockwise from top left: Vikram Iyer, Shyam Gollakota holding smartphone, Elyas Bayati, Arka Majumdar, Rajalakshmi Nandakumar

The research team, clockwise from top left: Vikram Iyer, Shyam Gollakota, Elyas Bayati, and Rajalakshmi Nandakumar

The team demonstrated its prototype can charge a smartphone at roughly the same speed as a standard USB cable. The laser charging system is capable of delivering two watts of power in a steady stream from a distance of 4.3 meters, or roughly 14 feet, to an area of roughly 15 square inches in size. The system can be modified to expand the charging area to more than three feet, like a tabletop, from a distance of 12 meters or nearly 40 feet.

The UW team’s laser-based system has several advantages over existing approaches to wireless power. Near-field magnetic induction, for example, is a commercially available technology that is safe and efficient, but also limited in range. Methods for harvesting ambient power from far field microwave sources, on the other hand, can handle a longer range, but are suitable only for ultra-low power applications requiring only a few microwatts of power. The researchers anticipate that their approach can be extended to charging a range of devices, including tablets, cameras, and desktop computers.

In addition to Gollakota, Majumdar, and Iyer, the team includes co-lead author Elyas Bayati, a Ph.D. student in electrical engineering who works in the UW Applied Physics Laboratory, and Allen School Ph.D. student Rajalakshmi Nandakumar.

Read the UW News release here, visit the project website here, and read related coverage in Wired, ForbesNew AtlasDigital TrendsGeekWire, The Engineer, International Business Times, The Daily Mail, R&D Magazine, and SlashGear.

Photo credits: Mark Stone/University of Washington


February 21, 2018

UW and Microsoft researchers achieve random access in large-scale DNA data storage

Two MISL members performing wet-lab experiments

Allen School Ph.D. student Lee Organick (foreground) and Microsoft researcher Yuan-Jyue Chen in the Molecular Information Systems Lab (Dennis Wise/University of Washington)

University of Washington and Microsoft researchers revealed today that they have taken a significant step forward in their quest to develop a DNA-based storage system for digital data. In a paper published in Nature Biotechnology, the members of the Molecular Information Systems Laboratory (MISL) describe the science behind their world record-setting achievement of 200 megabytes stored in synthetic DNA. They also present their system for random access — that is, the selective retrieval of individual data files encoded in more than 13 million DNA oligonucleotides. While this is not the first time researchers have achieved random access in DNA, the UW and Microsoft team have produced the first demonstration of random access at such a large scale.

One of the big advantages to DNA as a digital storage medium is its ability to store vast quantities of information, with a raw limit of one exabyte — equivalent to one billion gigabytes — per cubic millimeter. The data must be converted from digital 0s and 1s to the molecules of DNA: adenine, thymine, cytosine, and guanine. To restore the data to its digital form, the DNA is sequenced and the files decoded back to 0s and 1s. This process becomes more daunting as the amount of data increases — without the ability to perform random access, the entire dataset would have to be sequenced and decoded in bulk in order to find and retrieve specific files. In addition, the DNA synthesis and sequencing processes are error-prone, which can result in data loss.

MISL researchers addressed these problems by designing and validating an extensive library of primers for use in conjunction with polymerase chain reaction (PCR) to achieve random access. Before synthesizing the DNA containing data from a file, the researchers appended both ends of each DNA sequence with PCR primer targets from the primer library. They then used these primers later to select the desired strands through random access, and used a new algorithm designed to more efficiently decode and restore the data to its original, digital state.

“Our work reduces the effort, both in sequencing capacity and in processing, to completely recover information stored in DNA,” explained Microsoft Senior Researcher Sergey Yekhanin, who was instrumental in creating the codec and algorithms used to achieve the team’s results. “For the latter, we have devised new algorithms that are more tolerant to errors in writing and reading DNA sequences to minimize the effort in recovering this information.”

Snapshot of Ok Go band members splattered with multi-colored paint from music video

A snapshot from Ok Go’s music video for “This Too Shall Pass,” which the MISL team encoded in DNA

Using synthetic DNA supplied by Twist Bioscience, the MISL team encoded and successfully retrieved 35 distinct files ranging in size from 29 kilobytes to over 44 megabytes — amounting to a record-setting 200 megabytes of high-definition video, audio, images, and text. This represents a significant increase over the previous record of 22 megabytes set by researchers from Harvard Medical School and Technicolor Research & Innovation in Germany.

“The intersection of biotech and computer architecture is incredibly promising and we are excited to detail our results to the community,” said Allen School professor Luis Ceze, who co-leads the MISL. “Since this paper was submitted for publication we have reached over 400 megabytes, and we are still growing and learning more about large-scale DNA data storage.”

With this new milestone, MISL researchers have succeeded in demonstrating how DNA-based data storage — known to be significantly denser and more durable than existing digital storage technologies — can be practical, too. The UW and Microsoft team estimates its approach will scale to physically isolated pools of DNA containing several terabytes each. When dehydrated for storage, these pools of data would be several orders of magnitude denser than tape. And as the costs associated with DNA sequencing and synthesis continue to decline, the team foresees substantial activity devoted to the development of DNA-based data storage in future.

“DNA data storage is an incredibly exciting area, and it is great to see our progress recognized by such a reputable publication as Nature Biotechnology,” said Microsoft Senior Researcher Karin Strauss, co-leader of the MISL and an affiliate professor at the Allen School. “We are enthusiastic to continue working at the intersection of biotechnology and IT.”

It was this intersection that initially interested Allen School Ph.D. student Lee Organick, who performed many of the wet-lab experiments the team used to validate its approach. Having made the leap from undergraduate studies in molecular biology to computer science, she is enthusiastic about the potential impact of the MISL’s approach.

Georg Seelig, Luis Ceze, Karin Strauss

Left to right: Allen School and UW Electrical Engineering professor Georg Seelig, Allen School professor Luis Ceze, and Microsoft researcher Karin Strauss (Tara Brown Photography)

“We’re at a time when a lot of groundbreaking research will be done at the intersection of fields,” said Organick. “When I heard about this project it seemed a bit outlandish, but it captured my imagination.”

The makeup of the lab — which unites researchers from multiple disciplines and organizations — is another plus, in Organick’s view.

“Having worked with such a creative and diverse team of people for several years now, they’ve shown me that projects like this one are achievable,” she said. “And it’s just as exciting as it was the first day.”

The MISL draws upon the varied expertise of researchers from the Allen School, UW departments of electrical engineering and bioengineering, and Microsoft. In addition to Ceze, Strauss, Organick, and Yekhanin, contributors to the Nature Biotechnology paper include lab members Siena Dumas Ang, Yuan-Jyue Chen, Randolph Lopez, Konstantin Makarychev, Miklos Racz, Govinda Kamath, Parikshit Gopalan, Bichlien Nguyen, Christopher Takahashi, Sharon Newman, Hsing-Yeh Parker, Cyrus Rashtchian, Kendall Stewart, Gagan Gupta, Robert Carlson, John Mulligan, Douglas Carmean, and Georg Seelig.

Read the Nature Biotechnology paper here and related articles in IEEE Spectrum, ZDNet, and GeekWire.

Support the next phase of the team’s research by submitting an original image to the lab’s digital time capsule, which will be used to develop and refine techniques for processing visual information in DNA molecules, as part of the #MemoriesInDNA Project here.


February 19, 2018

Maya Cakmak named 2018 Sloan Research Fellow

Maya CakmakProfessor Maya Cakmak, director of the Allen School’s Human-Centered Robotics Lab, has earned a 2018 Sloan Research Fellowship from the Alfred P. Sloan Foundation. The fellowship recognizes Cakmak as one of the most outstanding young researchers in North America and a future scientific leader. Cakmak is among 16 computer scientists to receive a 2018 fellowship, out of 126 fellows overall drawn from more than 50 colleges and universities across the United States and Canada.

“The Sloan Research Fellows represent the very best science has to offer,” Sloan Foundation President Adam Falk said in a press release. “The brightest minds, tackling the hardest problems, and succeeding brilliantly — Fellows are quite literally the future of twenty-first century science.”

Cakmak’s research focuses on the development of general-purpose robots that can be programmed by their end-users, including people living with motor impairments, older adults who wish to age in place, and others who may require assistance in order to maintain their independence. According to Cakmak, it’s an idea whose time has come — one that is made possible by recent advances in the field.

“Within the last decade, developments in robotics, such as common hardware platforms and open-source software, have fueled a great deal of progress towards this vision,” Cakmak explained, citing examples of robots capable of performing household tasks such as folding laundry or emptying a dishwasher. “However, these capabilities are far from ready for the real world for two important reasons: they are tailored to a particular environment in which the robot operates, meaning that they can’t be easily adapted to other settings, and they involve long development cycles and programmers capable of using specialized software.”

The first problem, Cakmak says, is already being addressed through mainstream robotics research that aims to develop universal capabilities that will work correctly in every potential scenario. But those efforts are hampered by the difficulty for researchers of anticipating every variable a robot might encounter in the course of its work. For her part, Cakmak is interested in eliminating the first problem altogether by tackling the second: empowering users to program their own robots to suit their individual needs and environment. To that end, she is working on ways to enable anyone to do what currently can be done only by skilled programmers. Her approach includes teaching robots to physically manipulate objects through demonstration and verbal commands, and the development of new visual and textual robot programming languages that make programming robots simpler and faster than current software engineering practices.

Ultimately, Cakmak’s work is guided by her desire to have a real impact on real people.

“I deeply care about the relevance and usefulness of my research,” Cakmak said. “We try to evaluate systems we develop in my lab with realistic and diverse sets of tasks, putting them in front of potential users with diverse backgrounds and abilities, and we take every opportunity to demonstrate and deploy them in the real world.”

Cakmak and her students collaborate with robotics startups like Savioke and Fetch Robotics to develop tools that can be useful for these companies and their customers. She also runs a summer camp where high school students with diverse disabilities learn to program robots using tools developed in her lab.

Cakmak is the 32nd current or former Allen School faculty member to have earned one of these prestigious awards, for which candidates are evaluated by senior scientists based on their research accomplishments, creativity, and potential to be a leader in their respective fields. She joins recent recipients Ali Farhadi and Jon Froehlich (2017), Emina Torlak (2016) and Emily Fox, Shyam Gollakota, and Thomas Rothvoss (2015).

Cakmak is not the only computer scientist with an Allen School connection to be honored by the Sloan Foundation this year. Former postdoc Simon Peter, who worked with Allen School professors Tom Anderson and Arvind Krishnamurthy, also received a 2018 fellowship. Peter is a faculty member at the University of Texas at Austin, where he focuses on operating systems and networks.

Four other UW faculty members are among the current class of fellows: Jiun Haw-Chu of the Clean Energy Institute, Arka Majumdar of Electrical Engineering and Physics, Jessica Werk of Astronomy, and Chelsea Wood of Aquatic & Fishery Sciences.

Read the Sloan Foundation press release here and the UW News release here. View the complete list of 2018 Fellows here.

Congratulations, Maya and Simon!


February 16, 2018

Professor Michael Ernst honored for extraordinary contributions to student mentorship

Michael ErnstAllen School professor Michael Ernst has earned the 2018 CRA-E Undergraduate Research Faculty Mentoring Award from the Computing Research Association. The award recognizes faculty who provide exceptional mentorship and support to student researchers. Ernst is a member of the Allen School’s Programming Languages & Software Engineering (PLSE) group.

Recipients of the CRA-E mentorship award are chosen based on their track record of providing a high-quality, rewarding research experience to aspiring computer scientists. The CRA cited Ernst’s combination of research accomplishments and caring demeanor in its award announcement, which also praised his record of mentorship across multiple institutions, projects, and publications. To date, Ernst has mentored 123 undergraduate students over the course of his research career, co-authoring more than 50 publications with them along the way.

“He is typically described as a caring and careful mentor who is selfless, patient, quality-driven, and student-focused,” the CRA said of Ernst, noting that nearly half of his undergraduate mentees have gone on to attend graduate school in computer science.

Ernst’s success can be attributed in part to his holistic approach to student mentorship. In addition to providing direct guidance to mentees, he cultivates an environment and a team that is welcoming and supportive of young researchers — including instructing graduate students, postdoctoral researchers, and faculty colleagues on how to successfully work with undergraduates. He also incorporates state-of-the-art research tools and project-based exploration in his undergraduate classes.

“I’m passionate about mentoring because doing research as an undergraduate changed the course of my career and my life,” said Ernst. “I love making new discoveries — and I get a vicarious thrill from helping others to experience that same wonder.”

Ernst’s approach has made him a role model for the next generation of computer science faculty. As one former mentee who went on to become a faculty member said, “I strive to use as much of prof. Ernst’s mentoring style in my own advising as I can.” If imitation is the highest form of flattery, you can’t get much higher praise than that.

In addition to Ernst, the CRA recognized Catherine Putonti, a professor of computer science and biology at Loyola University Chicago and the Stritch School of Medicine, with a 2018 faculty mentoring award. Read the full CRA announcement here.

Congratulations, Mike — and thank you for delivering an exceptional research experience to so many Allen School undergraduates!


February 13, 2018

Allen School undergraduates earn national recognition for research excellence

Computing Research Association logoIn keeping with the Allen School’s commitment to provide an unparalleled educational experience to students, many undergraduates participate in leading-edge research in our labs. Undergraduate researchers work alongside faculty, postdocs, and graduate students and often submit their work to major academic research conferences and scientific journals. This year, three of these talented student researchers — Kimberly Ruth, Preston Jiang, and Deric Pang — were recognized as part of the Computing Research Association’s 2018 Outstanding Undergraduate Researcher Awards, which highlight exceptional potential among young computer scientists.

Kimberly RuthKimberly Ruth, a junior who is double-majoring in computer engineering and mathematics, was named a finalist in this year’s CRA awards competition. For the past two years, she has worked with professors Tadayoshi Kohno and Franziska Roesner in the Allen School’s Security and Privacy Research Lab as part of a team focused on security for emerging augmented reality platforms.

Ruth collaborated with Ph.D. student Kiron Lebeck on the development of Arya, an AR system that protects against buggy or malicious output by applications. As part of that work, Ruth contributed to the prototype built on the Unity game engine, including a selection of applications in C# to run on the system. The team’s results were published last year at the 38th IEEE Symposium on Security & Privacy, with Ruth as the second author, and subsequently invited for publication in an upcoming issue of IEEE Security & Privacy Magazine. She also worked with Lebeck to design and execute a user study to evaluate people’s reactions and concerns related to AR technology, with an emphasis on multi-user scenarios, using the Microsoft HoloLens. Inspired by the results of that user study, Ruth has taken the lead on a new project examining a set of open questions related to the design of multi-user AR platforms that span computer privacy and security, operating system design, and human-computer interaction.

Outside of the lab, Ruth volunteers as a peer tutor for the Allen School’s Foundations of Computing course. She previously earned a 2017 Mary Gates Research Scholarship, a competitive award that recognizes University of Washington students engaged in undergraduate research, and a 2017-18 Washington Research Foundation Fellowship, which supports students engaged in sophisticated science and engineering research projects.

Preston JiangPreston Jiang is a transfer student from Seattle Central College who works with Rajesh Rao, the Hwang Professor of Computer Science & Engineering and Electrical Engineering and director of both the Allen School’s Neural Systems Lab and the National Science Foundation’s Center for Sensorimotor Neural Engineering. He also works with professor Andrea Stocco, co-director of the Cognition & Cortical Dynamics Laboratory. Jiang was recognized with an Honorable Mention from CRA for his research on brain-computer interfaces (BCIs) and brain-to-brain interfaces (BBIs).

Jiang started out applying his knowledge of signal processing, machine learning and systems integration to create a new BCI for controlling a cursor based on electroencephalography (EEG) signals from the scalp. Rao subsequently tapped him to serve as the lead student researcher on the third generation of their BBI project, which seeks to enable individuals to communicate directly using EEG and signals transmitted to the visual cortex via a transcranial magnetic stimulator (TMS). Jiang’s contributions focused on BrainNet, a system for linking three or more brains to collaboratively solve a task through visual brain signals. As part of this work, he helped to assess the viability and performance of the BrainNet, managed integration of the EEG, BCI, and TMS systems, and ran experiments with human subjects. Jiang also contributed to another ambitious project, this time exploring the potential for brain-based virtual reality, by cataloguing the variety of artificial visual sensations known as “phosphenes” that humans experience to enable the future development of VR systems based on direct brain stimulation.

The team plans to publish journal papers on both BrainNet and the brain-based VR project with Jiang as co-author. Jiang previously was recognized for research excellence with a 2017 WRF Innovation Undergraduate Fellowship in Neuroengineering and a 2017-18 Levinson Emerging Scholars Award, which supports juniors and seniors pursuing advanced research in bioscience and related areas.

Deric PangDeric Pang is a student enrolled in the Allen School’s combined bachelor’s/master’s program who earned an Honorable Mention in the CRA awards competition. He has contributed to research projects on fault localization, programming by natural language, and software testing under the guidance of professor Michael Ernst in the Programming Languages & Software Engineering (PLSE) group.

Pang was part of a team that changed our understanding of fault localization, an approach for identifying defective lines of code to save substantial effort in the debugging process. Previous research into fault localization techniques has tended to use artificial faults for ease of evaluation. However, Pang and his colleagues discovered that a technique’s performance on artificial faults is not a good predictor of how well it will handle real faults. After replicating previous work, the team found that only 30 percent of the evaluated techniques were statistically and practically significant for artificial faults — and none met this threshold for real faults. Pang and his fellow researchers used their findings to develop new, more effective techniques and presented their work at the 39th International Conference of Software Engineering (ICSE 2017). Pang also contributed to Tellina, a code translation tool that is the result of a collaboration between PLSE and a group of researchers led by professor Luke Zettlemoyer of the Allen School’s Natural Language Processing group. Tellina translates the natural language description of a desired operation into programming language using recurrent neural networks. The system enables programmers to be more productive by allowing them to describe an intended operation using their own words rather than having to memorize the details of increasingly complex systems. In a controlled user study, Pang and his colleagues found that, even in cases where Tellina’s predictions were not completely correct, programmers who used the tool significantly outperformed those who did not.

Pang previously completed internships at Marchex and Amazon, where he worked on automatic speech recognition and the Alexa Machine Learning team, and studied abroad as a computer science exchange student at ETH Zürich. He plans to spend this spring as an intern at NVIDIA working on autonomous drone research.

Since 2000, a total of 58 Allen School students have been honored by CRA for excellence in undergraduate research. Congratulations to Kimberly, Preston, and Deric — and thanks to the faculty and graduate students who serve as mentors and enthusiastically support undergraduate research!


February 12, 2018

Yin Tat Lee wins NSF CAREER Award to develop new, efficient algorithms for convex optimization

Yin Tat Lee head shotProfessor Yin Tat Lee of the Allen School’s Theory of Computation research group has received a CAREER Award from the National Science Foundation to develop faster, more efficient algorithms for solving convex and other optimization problems. The outcome of Lee’s research, which seeks to increase the scientific community’s understanding of the relationship between convex geometry and optimization algorithms and improve upon current techniques drawn from continuous and discrete optimization, will have broad impact across the sciences and beyond.

Convex optimization techniques have applications in a range of fields, including machine learning, statistics, mathematics, economics, and operations research. However, many of these techniques historically tended to be inefficient and expensive to implement. Recent advances yielding faster algorithms have enabled Lee to break the long-standing running time barriers for specific problems, such as linear programming and maximum flow problems, and to apply optimization techniques to a broader class of problems than was previously feasible. Lee aims to build upon that past work by tackling a set of significant problems in convex geometry and optimization in order to push the state of the art even further. His approach will draw from techniques used in a variety of domains, including combinatorial and convex optimization, convex and Riemannian geometry, spectral graph theory, stochastic processes, and more.

One major goal of Lee’s CAREER proposal is to make progress towards resolving the Kannan-Lovasz-Simononoviz (KLS) conjecture — a central problem not only to the field of optimization but to theoretical computer science and mathematics, and one that implies several other well-known conjectures. Such progress would represent a significant breakthrough in researchers’ understanding of convex optimization and yield immediate running-time improvements for several problems in convex geometry. Lee also aims to resolve a number of algorithmic barriers to optimization, most notably the square-root iterations barrier for solving linear programs. This work would overcome a major obstacle to achieving nearly linear-time algorithms for the maximum flow problem, which is a key subroutine in many other algorithms and promises to have broad theoretical and practical implications.

Finally, Lee will apply a geometry perspective to the study of complex optimization algorithms, such as first-order methods and cutting-plane methods, in order to better understand their complexity and aid in the discovery of new applications. He also intends to explore the use of sampling algorithms for non-convex optimization, which is a rapidly developing area in machine learning.

The CAREER Award is administered through NSF’s Faculty Early Career Development Program and is designed to recognize and support promising junior faculty who successfully blend teaching and research and demonstrate the potential to be leaders in their respective fields. Including Lee, 59 Allen School faculty members have received one of these prestigious awards or their predecessor, the Presidential/NSF Young Investigator Award.

Read Lee’s award abstract here.

Congratulations, Yin Tat!


February 7, 2018

All work and no play …

The antidote to becoming a dull computer scientist: the Allen School’s annual ski day at Stevens Pass!

January 27, 2018

Deepali Aneja and Eric Whitmire named Adobe Research Fellows

Deepali Aneja

Allen School Ph.D. students Deepali Aneja and Eric Whitmire have been named 2018 Adobe Research Fellows. The Adobe Research Fellowship program recognizes outstanding graduate students with exceptional technical and personal leadership skills who are engaged in creative, high-impact research. Aneja and Whitmire are among a total of 10 graduate students around the world to be recognized in this year’s fellowship competition.

Aneja works with Allen School professor Linda Shapiro in the Graphics and Imaging Laboratory (GRAIL) and Creative Director Barbara Mones in the Animation Research Labs. Her research focuses on computer vision, the intersection of vision and machine learning, and computer graphics and animation.

Aneja previously spent two summers as an intern in Adobe Seattle’s Creative Intelligence Lab, where she contributed to the team’s efforts to enhance lip sync accuracy for 2D animation in Adobe Character Animator 2018, part of the company’s Creative Cloud lineup. Previously, she completed a computer vision research internship at Lawrence Livermore National Laboratory as part of its Data Heroes intern program. At last year’s Allen School open house and poster session, Aneja and her collaborators captured the People’s Choice Award for “Learning Stylized Character Expressions from Humans.” The winning poster presented the team’s work on DeepExpr, a data-driven framework that uses deep learning to map human facial expressions to animated characters in a way that is both perceptually valid and geometrically correct.

Eric Whitmire

Whitmire works with Allen School and Electrical Engineering professor Shwetak Patel in the UbiComp Lab, where his research focuses on the intersection of hardware and software to enable new capabilities in wearable sensors, interaction, and mixed-reality systems.

Whitmire has completed multiple internships at Oculus Research, where he worked on alternative input techniques for augmented reality applications and a wearable scleral coil eye-tracking system for virtual reality displays. He earned the Best Paper Award for his work on the latter, called EyeContact, at the Association for Computing Machinery’s International Symposium on Wearable Computers (ISWC) in 2016. Whitmire spent last summer working with a team at Microsoft Research on a new handheld controller with haptic feedback for virtual reality applications. Other recent projects include DigiTouch, a reconfigurable glove that enables thumb-to-figure touch interaction for general input and text entry on head-mounted AR and VR displays, and PupilScreen, a smartphone app currently under development that will enable accurate, on-the-spot assessment of traumatic brain injury.

The Adobe Research Fellowship comes with a financial award, an Adobe Research mentor, and the opportunity to spend a summer as an intern at Adobe. Past Allen School recipients include Ph.D. students Julian Michael, who works with professor Luke Zettlemoyer in the Natural Language Processing group, and Pavel Panchekha, who works with professors Michael Ernst and Zachary Tatlock in the Programming Languages & Software Engineering (PLSE) group.

Congratulations, Deepali and Eric!


January 25, 2018

#MemoriesInDNA Project aims to crowdsource 10,000 images to advance DNA-based data storage

What do you want to remember forever?

MISL researcher piping synthetic DNA

MISL researchers aim to collect 10,000 images to encode in synthetic DNA to develop new techniques for data storage, search, and retrieval.

That’s a question that researchers in the University of Washington’s Molecular Information Systems Laboratory (MISL) hope will inspire people around the world to submit original photos to the #MemoriesInDNA Project. The project — the result of a partnership between Allen School, UW Department of Electrical Engineering, Microsoft, and Twist Bioscience — aims to build a robust dataset of 10,000 images to develop exciting new capabilities for DNA-based data storage and processing.

The MISL launched in 2015 to develop synthetic DNA as an archival storage medium for digital data that is denser and more durable than existing technologies. Now, backed by a $6.3 million grant from the U.S. Defense Advanced Research Projects Agency (DARPA) as part of its Molecular Informatics program, MISL researchers plan to build upon their prior work. Using the trove of visual data that will be assembled as part of the #MemoriesInDNA Project, the team will explore new ways to process and search for data still encoded in DNA — without having to retrieve and convert the images back into their electronic form. It’s the next frontier in the evolution of DNA as a viable — and truly useful — solution for the world’s growing data storage needs.

“Let’s suppose you have a trillion images encoded in DNA and want to find all the photographs that have a red car in them,” Allen School professor Luis Ceze explained in a UW News release. “We want to be able to do that information processing in DNA directly — to search in a smart way and make the molecules themselves carry out that computer vision work.”

People around the world are invited to submit photos of people, places, and moments that they want to remember forever. Here is a sampling of images submitted via the upload site.

To achieve this “smart” search capability, Ceze and his colleagues will leverage the tendency of certain nucleotides that make up DNA molecules to bind themselves to others — adenine (A) to thymine (T), and cytosine (C) to guanine (G). As part of the encoding process, MISL researchers convert the digital data of an image — 0s and 1s — to the A, T, C, and G molecules that make up strands of DNA. To retrieve only those images they are interested in out of the thousands that make up the dataset, without having to convert them back to binary, the researchers plan to introduce a query containing complementary DNA that will cause only those that meet their search criteria to bind to it. The inclusion of magnetic nanoparticles in the query will enable them to pull out the images bound to it with the help of a magnet. The team will also employ machine learning techniques to enable the detailed mapping and encoding of all visual features that may be contained in an image to enable scientists to perform meaningful data processing.

The MISL team has already set a world record for the amount of digital data stored in and successfully retrieved from DNA, from the hip (a video by the band OK Go) to the historic (the Universal Declaration of Human Rights in 100 languages). To develop a robust capability to search digital data within the DNA itself, however, the team needs a significantly larger volume and variety of images to work with. That’s where the #MemoriesInDNA social media campaign, also launched today, comes in.

“It’s your turn to show us what should be preserved in DNA forever,” Ceze said. “We want people to go out and take a picture of something that they want the world to remember — it’s a fun opportunity to send a message to future generations and help our research in the process.”

The team plans to eventually make this digital time capsule — stripped of any personally identifying information — available to researchers around the world.

Ana Mari Cauce and Paul Allen onstage

The Allen School’s contribution to the #MemoriesInDNA digital time capsule: UW President Ana Mari Cauce and Paul G. Allen celebrating the naming of the Paul G. Allen School of Computer Science & Engineering on March 9, 2017.

“It is thrilling to bring computer science and molecular biology together in this project,” said Microsoft senior researcher Karin Strauss, an affiliate associate professor at the Allen School. “There has been amazing progress recently in both areas and, when combined, they can be very powerful in tackling problems created by the massive amounts of data we’ve been generating.”

Other lead contributors to the project include Allen School and Electrical Engineering professor Georg Seelig and Microsoft partner architect Douglas Carmean. Twist Bioscience will supply the synthetic DNA for the project.

Snap a photo for science!

Anyone can contribute to the data set by uploading an original photo via the website Afterward, help the campaign go global and inspire others to participate by sharing your image on social media with the hashtag #MemoriesInDNA.

Read the UW News release here , a related post on the Twist Bioscience blog here, and the Wired article here.


January 24, 2018

Alumna Gail Murphy receives Harlan D. Mills Award for contributions to software engineering

Gail MurphyAllen School alumna Gail Murphy (Ph.D., ’96), professor of computer science and Vice President Research & Innovation at the University of British Columbia, has been recognized by the IEEE Computer Society with its Harlan D. Mills Award, which honors individuals for significant contributions to software engineering research and practice through the development and application of sound theory. Murphy earned the 2018 award for “outstanding research on understanding software-development practices and tools that improve the productivity of developers.”

Murphy’s research focuses on the development and evaluation of tools for to enable software developers to better identify, manage, and coordinate the information that matters most to their work. She also conducts empirical studies to better understand the software development process. Her focus was partly inspired by experience: Murphy started out as a senior software engineer at MPR Teltech before arriving at UW to earn her Ph.D. with the late David Notkin. She later co-founded Tasktop Technologies to commercialize this work and streamline software development at the enterprise level.

The Harlan D. Mills Award is the latest in a string of professional honors Murphy has collected over the years, including the AITO Dahl-Nygaard Junior Prize, which highlights promising contributions to the field of object-oriented programming, and the CRA-W Anita Borg Early Career Award in recognition of her professional contributions and her commitment to increasing the participation of women in computing. Murphy is a Fellow of the Royal Society of Canada and of the Association for Computing Machinery, and she has earned a number of Distinguished Paper and Most Influential Paper awards — a testament to the enduring impact of her work in the software engineering community and the field of computer science, generally. In addition to professional accolades, Murphy has also earned multiple honors from her alma mater, including the Allen School’s 2014 Alumni Achievement Award and a 2008 Diamond Award from the UW College of Engineering for early-career achievement.

Read the full Harlan D. Mills Award announcement here.

Congratulations, Gail!


January 22, 2018

« Newer PostsOlder Posts »