Students explore sustainable building with bamboo

Watch the video

LAST FALL, two undergraduates approached Sigrid Adriaenssens, an associate professor of civil and environmental engineering, about working together on their senior-thesis projects, from different angles.

Lu Lu, who is from Chongqing, China, wanted to work on sustainable construction with a focus on design and digital modeling. Russell Archer, who is from East Orange, New Jersey, wanted to physically test building materials. Adriaenssens served as the adviser for both students, who graduated in 2016.

With the help of a graduate student in Adriaenssens’ lab, the seniors identified a partner, the Administrative Department of Environmental Management (DAGMA), in Cali, Colombia. The students and the group collaborated on a project involving bamboo architecture and construction — the entrance canopy to a park to be used by schoolchildren.

Bamboo grows quickly and is lightweight but strong. It has been used as a building material for centuries, but little engineering analysis has been done on it. Lu focused on the structural form of the canopy, and Archer analyzed the effectiveness of the fishmouth joints used in the designs.

Last March, with funding from the School of Engineering and Applied Science, they traveled to Cali to meet with the DAGMA architects and engineers, share their results, and inform the design and construction of the bamboo canopy.

“I couldn’t have imagined I would be traveling to Colombia,” said Archer, who plans to pursue a master’s degree in structural engineering. “I’m really impressed by the broadness of the entire project.”

“I really enjoyed this project because engineering is not only hard-core science — you calculate something but that’s it,” Lu said. “There’s a very strong social component.”

–By the Office of Communications

Exploring collective interactions of matter and antimatter

STRIP AWAY ELECTRONS FROM THEIR ATOMS and you get a plasma — a collection of negatively charged electrons and positively charged ions. But at high energies around compact cosmic objects such as black holes, quasars and pulsars, curious plasmas may form that, instead of ions, contain positrons, the antimatter counterparts of electrons.

Scientists are searching for ways of distinguishing this type of plasma from others, both in astrophysical environments and in laboratories on Earth. Julia Mikhailova, an assistant professor of mechanical and aerospace engineering, and Matthew Edwards, a graduate student in her lab, together with Professor of Astrophysical Sciences Nathaniel Fisch, found that, contrary to earlier claims, an electron-positron plasma would scatter some wavelengths of light surprisingly intensely via a process called Brillouin scattering.

This fundamental insight into the unusual behavior of matter-antimatter plasmas, published in the journal Physical Review Letters Jan. 8, 2016, may help to find such plasmas in space, or validate methods for creating them in the lab. The work was funded in part by the National Science Foundation and the National Nuclear Security Administration. –By Bennett McIntosh

Researcher probes the secret life of electrons

ELECTRONS DART within and between atoms far too quickly for current imaging techniques to observe their motion. To capture fast-moving objects without a blur, a photographer can use a camera flash to light up a scene for an instant. Julia Mikhailova, an assistant professor of mechanical  and aerospace engineering, hopes to capture electron motion in a similar way, but her camera flash must last only a few attoseconds — just millionths of a trillionth of a second.

Mikhailova and her team use lasers and plasmas, which are collections of charged particles, to create attosecond pulses of light. “With these pulses one can observe the action inside atoms and molecules,” Mikhailova said. These observations could help researchers predict electron behavior, leading to a better understanding of everything from chemical reactions to superconductivity.

Lasers cannot by themselves produce such brief pulses. Instead, scientists scatter high-powered laser light off a stream of gas. Too much energy from the laser, however, can strip the electrons in the gas from their atoms to make a plasma that no longer scatters light in the right way.

Mikhailova and her team, with funding from the National Science Foundation, are taking a different approach. Instead of using a gas target, they aim a much higher-powered laser at a solid glass disc, creating a dense plasma at its surface. The laser light — an oscillating electromagnetic field — accelerates the electrons in the plasma toward nearly the speed of light. At these speeds, Newton’s physics breaks down and relativity takes over, causing the release of light in the form of attosecond pulses.

To improve the technique, Matthew Edwards, a graduate student in Mikhailova’s lab, ran simulations of plasma-laser interactions on Princeton’s TIGRESS high-performance computer cluster. The results, published Sept. 16, 2016, in the journal Physical Review Letters, showed that mixing laser light with light at harmonic frequencies — which are multiples of the original laser’s frequency — increases the efficiency of the process.

In a paper published Feb. 24, 2016, in the journal Physical Review A, the researchers proposed an extremely efficient two-target system: the first target generates light with harmonic frequencies, which hits the second target to generate attosecond pulses. With such powerful pulses, a new domain becomes visible, Mikhailova said. “We may be able to really see what the electron is doing as it ‘orbits’ in the atom.” –By Bennett McIntosh

Bias in the machine: Internet algorithms reinforce harmful stereotypes

THE ARTIFICIAL-INTELLIGENCE (AI) SYSTEMS that suggest our search terms and otherwise determine what we see online rely on data that can be biased against women and racial and religious groups, according to a study led by researchers in Princeton’s Center for Information Technology and Policy (CITP).

As machine learning and AI algorithms become more ubiquitous, this phenomenon could inadvertently cement and amplify bias that is already present in our society or a user’s mind, according to the study, which was led by Arvind Narayanan, an assistant professor of computer science. The paper was posted in August 2016 on the preprint server arXiv.

The team found that the algorithms tended to associate domestic words more with women than men, and associated negative terms with the elderly and certain races and religions. “For just about every kind of bias that’s been documented in people, including gender stereotypes and racial prejudice, we were able to replicate it in today’s machine-learning models,” said Narayanan, who worked with CITP postdoctoral research associate Aylin Caliskan-Islam and Joanna Bryson, a professor of computer science at the University of Bath and a visiting scholar at CITP.

Machine-learning algorithms build models of language by exploring how words are used in context — for example, by combing all of Wikipedia or gigabytes of news clippings. Each time the model learns a word, it gives that word a series of geometric coordinates that correspond to a position in a many-dimensional constellation of words. Words that are frequently found near each other are given nearby coordinates, and the positions reflect the words’ meanings.

Biases develop as a result of the positions of these words. If the text used to train the model Bias in the machine: Internet algorithms reinforce harmful stereotypes more often associates “doctor” with words relating to men, ambition and medicine, while linking “nurse” to words related to women, nurturing and medicine, the model would come to assume that “nurse” is feminine, possibly even the feminine version of the masculine “doctor.”

To measure biases in algorithm results, the researchers adapted a test long used to reveal implicit bias in human subjects, the Implicit Association Test, for use on the language models. The human version of the test measures how long it takes a subject to associate words such as “evil” or “beautiful” with names and faces of people from different demographics. Thanks to the geometric model of language that machine-learning algorithms use, their biases can actually be measured more directly by simply finding the distance between the name of a group and positive, negative or stereotypical words.

Such biases can have very real effects. For example, in 2013 researchers at Harvard University led by Latanya Sweeney noted that African American-sounding names were far more likely to be paired with ads for arrest records. Such experiences could lead to unintentional discrimination when, say, a potential employer searches the internet for an applicant’s name.

“AI is no better and no worse than we are,” Bryson said. “However, we can continue to learn, but the machine learning for an AI program might be turned off, freezing it in a prejudiced state.” If we can measure this bias, however, Narayanan said, we can take steps to mitigate it. This could mean mathematically correcting a language model’s bias or simply being aware of the algorithms’ faults — and our own. –By Bennett McIntosh

CITIES: Resilient • Adaptable • Livable • Smart

Innovations in building materials, design, water systems and power grids are helping to make cities more livable, say researchers in Princeton’s School of Engineering and Applied Science

By Bennett McIntosh

Cities. They sprawl and tangle, juxtaposing ancient public squares and glistening skyscrapers. They provide homes for half of humanity, and economic and cultural centers for the rest.

It has taken us thousands of years to build today’s urban centers, and yet, they’re expected to double in land-area in just the next few decades. “Half the urban infrastructure we will be using in 2050 has not yet been built,” said Elie Bou-Zeid, a Princeton associate professor of civil and environmental engineering.

Though this growth is inevitable, the way these cities will expand is not. Rather than repeat the sprawling and uncoordinated development patterns of the past, researchers like Bou-Zeid and others in Princeton’s School of Engineering and Applied Science are exploring new ways to build urban infrastructures to serve our growing population, changing civilization and warming planet.

These intelligent cities will require buildings that heat and cool themselves on a limited energy budget. They’ll require bridges and other infrastructure built with the flexibility to adapt to a changing global climate and rising sea levels. And they’ll require innovations in the networks that supply cities with water and energy. These ideas — from new building materials to continent-spanning electrical grids — have the potential to shift urban development away from the present-day jumble of strip malls, suburbs and shantytowns toward the resilient cities of the future.

Clever buildings

The basic unit of these smarter, resilient cities is the intelligent building. Assistant Professor Forrest Meggers, who has a background in architecture and engineering, has a number of plans for making buildings smarter about how they heat and cool their indoor spaces. Often these heating and cooling systems involve water, which readily absorbs heat that is then shed through evaporation.

In one structure called the Thermoheliodome, the interior is coated with mirrors at odd angles to reflect heat toward water-cooled pipes. In another, the interior cools itself with evaporation through an external membrane that traps liquid water while allowing water vapor to escape. By demonstrating the effectiveness of these innovative ideas, Meggers, who has a joint appointment in the School of Architecture and the Andlinger Center for Energy and the Environment, hopes to show other architects that it is possible to make more effective and more attractive heating and cooling systems.

Meggers’ structures take advantage of two different ways heat is transferred: It can be carried by molecules of warm air or water, or it can radiate like light directly from surface to surface. Thermometers, which measure air temperature, don’t capture the effects that radiative heating and cooling can have on a building’s occupants, so Meggers developed a radiative heat-sensing camera. About the size of a thermostat, the camera captures a 360-degree view that researchers can use to build a 3-D model of the radiative surfaces in any room.

To investigate urban radiant-heat exchanges, Meggers’ students took similar devices to New York City, about 50 miles northeast of Princeton. The resulting thermal photographs enabled them to see how heat lingers in alleyways and clusters around window-mounted air conditioners. By seeing the heat, architects and engineers can improve their designs for optimal energy efficiency.

Optimal cooling is the goal of one project that Meggers collaborated on with Dorit Aviv, who earned her master’s degree in architecture in 2013 and is now a doctoral student at Princeton. The building is called the Cool Oculus and is designed to keep cool in the desert heat through a combination of evaporation and shifting shape. The researchers built the Cool Oculus as a prototype on the Princeton campus, and have secured a grant from the New York-based Tides Foundation to build a full-scale model and measure its capabilities.

During a hot day, mist flows into the Oculus’ central chimney and evaporates to cool the air within, which sinks as a refreshing breeze into the building. Meanwhile, the structure’s foundation absorbs excess heat, which it releases at night when the chimney widens to expose the foundation to the cool night sky. Combined, these effects can turn 100-degree desert heat into a comfortable 75 degrees.

Inspired by nature

The Oculus moves on a daily cycle, but Sigrid Adriaenssens, an associate professor of civil and environmental engineering, has designed structures whose real-time response to heat is built into the material itself. In a transparent case above her desk, Adriaenssens displays three structures that could pass for the leaves of a cyborg Venus flytrap. They each are made of white translucent shells curving off a central metallic strut.

The resemblance to a flytrap is not coincidental. Adriaenssens designed the structures with inspiration from the waterwheel plant, an aquatic cousin of the flytrap. This shape allows the entire structure to open or close in response to a small movement of the central strut. The strut, in turn, is made of two metals that expand differently when heated, so the shell expands significantly with a small increase in temperature. Adriaenssens envisions that these shells, which were developed with funding from the Andlinger Center for Energy and the Environment, could cover a building’s entire façade. On hot days, the shells would expand and block heat from streaming in through the windows.

Structures like these, which make clever use of materials and their form, will be the key to affordable and efficient structures in future cities, according to Adriaenssens. To make these structures, Adriaenssens and her team use computer simulations to calculate the structure’s optimal form. Like optimized forms in nature, Adriaenssens’ structures often show striking curves, from spiraling earthen garden walls and arching steel footbridges, to shell-shaped pavilions with slats to keep out direct sunlight while allowing in scattered light and breezes. Nor are the spreading leaves Adriaenssens’ only dynamic structure: To protect coastlines from storm surges while keeping them visually uncluttered, she has designed thick elastic spherical membranes that will inflate and press together to hold off the waves.

Water world

Where does the water that surges around Adriaenssens’ barriers go? Where does the water that cools Meggers’ buildings come from? To plan something as complex as a future city’s water system requires not just understanding the interactions between structures like these, but understanding how the structures and the people affect each other. Such an undertaking requires cooperation between researchers from many different fields, and an understanding of the successes and failings of many different cities, according to Bou-Zeid.

Bou-Zeid first grew interested in cities when he was a mechanical engineering undergraduate at the American University of Beirut in Lebanon. “I thought I would be designing racecars or airplanes, but environmental problems that involve the interaction of humans with their surroundings are more interesting,” he said. During his graduate and postdoctoral studies, Bou-Zeid investigated how cities — with their skyscraper-created wind canyons and their innumerable sources of heat and steam — fundamentally alter the movement of air around them.

Bou-Zeid is interested in how this airflow affects an invisible but critical part of cities’ water systems: evaporation. Before a city is built, water evaporates out of plants and earth, cooling the area. But in built-up areas, dark asphalt absorbs heat. Water flows off impermeable pavement into storm systems before it has the chance to evaporate and take heat away with it, trapping heat in the buildings and the streets. This trapped heat can warm cities by 10 to 15 degrees Fahrenheit higher than the surrounding countryside. The so-called urban heat island raises energy consumption and contributes to climate change as we burn fossil fuels to cool ourselves.

Parks, greenbelts and green roofs covered in plants can solve this problem by encouraging cooling through evaporation, Bou-Zeid said. But it is not as simple as planting trees: While Baltimore’s greenbelts have cooled it significantly, drier cities like Denver and Phoenix may be better off saving water by cooling with traditional air conditioning. “How do you compare the value of a gallon of water and a kilowatt-hour of energy in different cities?” Bou-Zeid asked.

Bou-Zeid’s attempts to answer this question, and similar studies by other researchers in every aspect of the water cycle, led to the formation of the Urban Water Innovation Network. The Network, supported by a five-year grant from the National Science Foundation, includes engineers, architects and social scientists from 14 institutions who are studying how six American cities interact with water. Bou-Zeid, Princeton’s team lead for the network, is working with colleagues at the University of Maryland and Arizona State University to create software that will model everything water can do in a city. Such software could be used to predict the benefit of new water projects while accounting for local climate and geology.

The wildest possible experiments

To ensure that the urban landscape is accurately represented in such simulations, professor James Smith is leading a team of researchers from five universities in the network to produce extremely accurate maps of the rainfall and flooding in each of the cities. For Smith, the William and Edna Macaleer Professor of Engineering and Applied Science and professor of civil and environmental engineering, such studies of real cities are the only way to understand urbanization’s present and future effects.

“In cities,” he said, “the wildest possible experiments are being carried out for you.” Rivers are rerouted. Vast tracts of land are paved over. Artificial shorelines and skylines change the flow of water and air. It is up to researchers to watch and learn from these unprecedented alterations to the land and environment.

Collaboration within the network leads in surprising directions. Meggers, also a member of the network, found a way to combine his interest in efficient heat transfer with the water systems. With Sybil Sharvelle, a professor at Colorado State University, he is designing wastewater systems that recapture the heat from showers and other uses of hot water.

When the project ends in 2020, the network will release a report detailing its findings and recommendations for the cities under study. The research covers an environmentally diverse collection of cities so that the suggestions can be useful to cities across the country and, in some cases, around the world. In the meantime, the network connects researchers and government officials to craft individual recommendations on short-term projects. “We ask the policymakers what they need to know, and try to understand their constraints so that our recommendations can be implemented,” Bou-Zeid said.

It’s not the first time Bou-Zeid has worked to make small, efficient changes to cities. Simply painting black roofs white so that they reflect more light keeps buildings cooler and saves energy and money.

New York City has implemented this idea via their °CoolRoofs program, through which thousands of volunteers have painted roofs white since 2009. These efforts provided Bou-Zeid with more data than he could ever have achieved in a laboratory. He is using data from this experiment in conjunction with his models of urban air and heat flow to determine the cost and energy savings of painting roofs white.

Networks and grids

Painting roofs white is a relatively easy modification to make to a city, but other modifications require a new way of thinking. Our cities are already in need of upgrades to electricity supply and delivery systems. Going forward, our electricity will increasingly come from renewable sources such as solar and wind power, which, while better for the environment, can vary due to wind shifts and cloud cover.

With renewable energy making up only about 10 percent of power production in the United States, this variability is not yet an issue, said Warren Powell, a professor of operations research and financial engineering who studies networks such as electrical grids and transportation systems. “But I see us hitting problems at about 20 percent renewables,” Powell said.

This variability makes it hard to fully replace coal, the traditional workhorse of electricity generation, and natural-gas turbines, which can be ramped up quickly. “When the dust clears in 40 years, we’re still going to have some fossil energy,” Powell said. While large, efficient batteries could store wind and solar power and release it as needed, the marginal cost of battery storage increases as more batteries are added to the grid. “It is going to be hard to fight this curve,” he said.

Changes in how the power grid operates could help. Powell recently began a project in Brazil, where a drought has cut into Brazil’s heavy dependence on hydroelectric power. Powell has begun working with a group of Brazilian power companies to study strategies for managing the variability from the influx of wind power. Because of wind’s variability, this is not simply a matter of replacing one power source with another. Instead, Powell will be supervising the development of Brazil’s first grid model that can closely simulate the variability of wind. This model will be used to develop robust management policies and energy portfolios that would help Brazil optimize an energy system that depends heavily on  wind and solar.

New technologies deployed smartly will help, Powell said. For example, self-driving electrical vehicles can decrease congestion in dense cities and lend their batteries to the electrical grid, selling power when the city needs it most and recharging overnight from the grid’s excess capacity.

Ultimately, these changes in technologies and policy must work within the economic and social constraints of existing cities. Failing to understand and anticipate urban changes and growth leads to not just bad policy, but unenforceable policy, Bou-Zeid said. If a city tries to prevent urban growth, for example, by limiting new housing, the city will often still grow, but in unregulated and unhealthy shantytowns on the periphery. “You must accept urban expansion — you have to work with it,” Bou-Zeid said.

But the size and inertia of cities is an opportunity, too, Meggers said. “Cities have the power to make a change.”

If researchers and policymakers at Princeton and in cities around the world can collaborate, making clever use of form, physics and interacting components as a part of urban planning, then that change will be a positive one.

Engineering health solutions for all

Benjamin Tien


Hundreds of women die every day due to excessive bleeding after childbirth, but this can be prevented by an injection of the hormone oxytocin, which stimulates uterine contractions that reduce bleeding. Yet oxytocin must be kept cold, and developing countries often lack the resources to transport and store the hormone. Nor do they have enough trained personnel to administer the shot.

Benjamin Tien, who graduated from Princeton in 2015 with a degree in chemical and biological engineering, has been awarded a Fulbright grant to join a research team developing an aerosolized, inhalable version of oxytocin. He’ll work at the Monash Institute of Pharmaceutical Sciences in Australia to use a technique called “spray-drying” to turn liquid oxytocin into a dry powder that can be released from an inexpensive, disposable device. This would cut out the need for special storage conditions, eliminate the risk of infection from needles, and make the treatment available even to women giving birth at home.

While at Princeton, Tien conducted senior thesis research on nanoparticles for dealing with bacterial infections such as cholera or pseudomonas with Robert Prud’homme, a professor of chemical and biological engineering and Tien’s thesis adviser. During the summer before his senior year, Tien worked at a startup company in Boston called Diagnostics for All that develops medical diagnostics for developing countries.

While doing research in Australia, Tien will also work with the Poche Center for Indigenous Health, an organization that aims to improve the health of indigenous Australians with an emphasis on forming lasting relationships with the communities. “I want to get a better sense of the challenges people are facing on the ground, which will help me know what to focus on in my career,” Tien said.

-By Takim Williams

Tiny delivery capsules for new drugs

‘Jack’ Hoang Lu researches nanoparticles for drug delivery

Graduate student ‘Jack’ Hoang Lu works on engineering nanoparticles for targeted drug delivery and diagnostics in the laboratory of Robert Prud’homme, professor of chemical and biological engineering. PHOTO BY CATHERINE ZANDONELLA

Some drugs cannot be delivered via a normal pill or injection because they cannot readily dissolve in water. About 40 percent of new pharmaceuticals have this hydrophobic (water-fearing) character, and like a globule of oil in water, they are unable to reach their targets in the body.

Robert Prud’homme, professor of chemical and biological engineering, addresses this problem by putting the drugs inside of nanoparticles, each about one-thousandth of the width of a human hair, which are then covered with a polymer called polyethylene glycol. The small size enables dissolution of the drug to increase their bioavailability.

Nanoparticles can also tackle a different delivery challenge presented by a growing class of drugs called biologics that, as the name implies, are made from cellular or genetic components. The problem is that they are all water-soluble, which make them easy prey for patrolling proteins that identify them as foreign and degrade them. The solution is similar: nanoparticles protect the biologics long enough for them to carry out their mission.

In addition to increasing the time that drugs spend flowing through the body, Prud’homme’s nanoparticles are capable of targeted delivery. This is necessary in the case of toxic drugs, including many cancer treatments, which would damage healthy cells if allowed to roam freely throughout the body. Prud’homme puts appendages, called ligands, on the outside of his nanoparticles that, due to their specific shape and chemical properties, attach to their target and only to their target. In collaboration with Patrick Sinko at Rutgers University, Prud’homme’s group discovered the optimum density of an appendage called a mannose ligand, used to target cells harboring tuberculosis microbes. The unexpected result was that attaching more appendages to a nanoparticle decreased its targeting effectiveness. This demonstration of the complex interplay between engineered nanoparticles and how the human body responds to them is a major theme of the Prud’homme research team. In addition to tuberculosis, the researchers are using these principles to attack cancer, inflammation, and bacterial infections.

The project receives funding from the National Institutes of Health, the National Science Foundation, Princeton’s IP Accelerator Fund and the School of Engineering and Applied Sciences Old Guard Grant.

–By Takim Williams

Computer chip for point-of-care diagnosis

Lab on a chip is smaller than a penny.

Kaushik Sengupta and his team are developing a computer chip-based diagnostic system, which is smaller than a penny but contains hundreds of different sensors for simultaneous detection of disease-causing agents.

Assistant Professor of Electrical Engineering Kaushik Sengupta and his team are developing a computer chip-based diagnostic system, which rests comfortably on a fingertip but contains hundreds of different sensors for simultaneous detection of disease-causing agents. The eventual goal is to use the chip in a handheld, portable diagnostic device that could be deployed in health clinics around the globe, especially in resource-limited settings.

The chip detects and measures the presence of DNA or proteins to help diagnose health conditions. Most existing methods for detecting these agents involve shining light on fluorescent labels attached to the DNA or protein and reading a resulting signal. However, in many types of tests, the signal is so weak that complex optical equipment is necessary to read the signal.

To perform this analysis using a simple handheld device, Sengupta is co-opting silicon chip technology similar to that found in personal computers and mobile phones. “This is a great technology for handheld medical diagnostic devices because it allows us integrate extremely complex systems in a single chip at very low cost. The vision is to unleash Moore’s law in diagnostics,” said Sengupta, referring to Intel co-founder Gordon Moore’s observation that processing power in computer chips has increased rapidly over the years.

The team starts with highly sensitive light-detecting components, or photodetectors, that are already ubiquitous in smartphone cameras, then adds new optical processor capabilities to the chip. The researchers found a way to re-wire the architecture of the chip so that in addition to carrying electrical information necessary for image processing, the chip also interacts with the incoming photons from the fluorescent light, and can block them out, allowing the signal that carries information about the test sample to be detected and processed.

Lingyu Hong and Kaushik Sengupta

Lingyu Hong, a graduate student in electrical engineering, and Kaushik Sengupta, an assistant professor of electrical engineering at Princeton University are developing technology for use in a handheld diagnostic system for healthcare in resource-limited settings.

This ability to integrate optical elements with electronics inside a single silicon chip is enabling the team to build detection systems for both genetic material and proteins. Millions of photodetectors can already be crammed into smartphone cameras and Sengupta plans to put hundreds or even thousands of such sensors on the new chip to create a platform capable of testing many agents at once. In addition to being cheap and robust, this “lab-on-a-chip” will be user-friendly. Sengupta and his colleagues envision that the chips will be used in a portable device similar to a smartphone that can use an app to analyze the fluorescence data and display diagnosis results in a clear, simple format.

To make the device truly portable, it will be necessary to develop a small and lightweight apparatus to isolate proteins and genetic material from blood or other fluids, and Sengupta and his collaborators are working on this challenge. “The entire end-to-end system may take another couple of years to reach, but we’ve demonstrated the feasibility of the approach,” said Sengupta, who collaborates with Professor of Chemistry Haw Yang. “Princeton provides the kind of environment that makes it easy to reach out to faculty members across the campus and to work on creative endeavors that cut across traditional disciplines.”

The initial work on the chip was supported by Project X, a fund established through a donation from G. Lynn Shostack S’69 for the support of exploratory research. The project involvesgraduate students Lingyu Hong in the Department of Electrical Engineering and Hao Li in the Department of Chemistry, Postdoctoral Research Associate Simon McManus and undergraduate Victor Ying. Lingyu and Hao were awarded a Qualcomm Innovation Fellowship for 2015-16 for this work.

-By Takim Williams

The Hub: A new center opens its doors … to student entrepreneurship

The Hub


THE SOCIAL CAMPUS NETWORKING startup Friendsy began with a single campus network at Princeton and has since expanded to 230 campuses nationwide.

This June, Friendsy was one of the first startups to move into the University’s Entrepreneurial Hub, a new incubator space for faculty, students and alumni. Located in downtown Princeton, the Hub houses the Keller Center’s annual eLab Summer Accelerator Program — a launch pad for student startups — as well as an eLab Incubator program that enables students to pursue their entrepreneurial ambitions during the academic year. The Hub also offers shared working space for startups founded by faculty, students and alumni, and serves as the center for Princeton’s entrepreneurship education programs.

“eLab has been tremendously helpful for our growth as a startup company,” said Michael Pinsky, one of Friendsy’s founders, who graduated from Princeton this year with a degree in psychology. “The ability to work with mentors in the field and use the working space at the Hub with what became a very large Friendsy team was incredibly valuable, and it is undoubtedly a major reason for our success.”

Now in its fourth year, the eLab program runs for 10 weeks in the summer and culminates in Demo Days, held in New York City and Princeton, at which the teams present their work to entrepreneurs, investors and innovators. The program has become an integral part of the University’s effort to assist students and faculty with pursuing new ventures.

The Hub was established in response to recommendations made by a committee set up to explore ways to expand entrepreneurship education and help University students, faculty and alumni advance their creative ideas and make important contributions to society.

Elab Summer Accelerator Program

Seven teams of students participated in the eLab Summer Accelerator Program at Princeton’s Entrepreneurial Hub this summer. From left to right: Kehinde Ope, a student at the University of Delaware, and Achille Tenkiang, a member of the Class of 2017 at Princeton University, have formed a startup called BLOC with Saidah Bishop, a student at Dartmouth College (fourth from left). Third from left is Diogo Adrados, Princeton Class of 2015, from the startup Rodeo. On the far right, Michael Pinsky, Princeton Class of 2015, is a co-founder of Friendsy.  PHOTO BY JILL FELDMAN

Mung Chiang, who chaired the committee, said the new space provides “an essential anchor” for a wide range of entrepreneurial activities at Princeton. The 10,000-squarefoot facility offers meeting rooms, offices and information technology support for startups sharing the co-working space. The Hub is also the home of the Princeton Entrepreneurship Council, led by Chiang and established in July 2015 to coordinate entrepreneurship programs on campus.

“The University has taken an important initiative in creating space for entrepreneurs and entrepreneurship education,” said Chiang, the Arthur LeGrand Doty Professor of Electrical Engineering and director of the Keller Center.

Celebrating its 10th anniversary in 2015, the Keller Center’s mission is to educate students as leaders in a technology-driven society by innovating education and fostering entrepreneurship, creativity and design. The center bridges disciplines to ensure that all students are prepared to put science and technology to use in solving critical societal challenges.–By John Sullivan

This year’s eLab teams include:
BLOC logoBLOC An online professional network for black collegians on the rise

clickstick Logo 6ClickStick Innovative dispensing technology with accurate dosage control for personal care, cosmetic and pharmaceutical products

Bodhi TreeBodhi Tree Systems An enterprise software system that facilitates the design and management of pharmaceutical trials

Friendsy LogoFriendsy A college-based social networking service that promotes friendships and relationships among members

KLOSKLOS Guitars A durable, affordable and comfortable carbon-fiber travel guitar

RodeoRodeo A mobile platform for users to browse and discover live events in their community

TeachMe_logoTeachMe A platform that connects college students to share knowledge, skills and experience with others in their community

Download PDF

Taming the network: Finding relationships in complex data sets

WHAT BRINGS PEOPLE TOGETHER IN ONLINE NETWORKS? Researchers (and advertisers) would like to know, but without access to personal profiles, the question is not easy. Finding previously undetected relationships in networks and complex data sets is one of the major challenges in the age of “big data.”

Now Assistant Professor Emmanuel Abbe and his collaborators have come up with a new way of thinking about networks to accomplish this task. Not limited to exploring social communities, the technique can tackle significant challenges such as determining which genes work together to increase your cancer risk or how to identify objects — such as chairs or puppies — in a collection of digital images.

The method involves examining whether members of a network are connected by looking at how many common “friends” they have, how many common friends those friends have, and so on. Using this information, the researchers construct a set of statistics that can predict who is in the same sphere.

The approach extracts the “signals” of communities amid a background of “noisy” connections. Abbe’s method is analogous to work by Claude Shannon, sometimes called the father of information theory, who showed that noise imposes a limit to the rate at which data can be transmitted with almost zero error. Abbe has shown that there is an analogous limit to the problem of recovering communities from large data sets.

“Once we understood that there is a fundamental limit to this problem, there was a clear line of sight for how to solve it,” said Abbe, a member of Princeton’s Department of Electrical Engineering and Program in Applied and Computational Mathematics. Abbe and Colin Sandon, a graduate student in the Department of Mathematics, put the method to the test by examining political blogs, some right-leaning and others left-leaning, that sprung up prior to the 2004 presidential election. They asked, if you knew which blogs were referring to each other, but had zero information about the content of the blog, could you figure out which blogs are run by Republicans and which ones by Democrats? “We were able to identify 95 percent of the blogs that we looked at as left- or right-leaning,” Abbe said.

The work was published in the Proceedings of the Annual Symposium on Foundations of Computer Science in 2015. Abbe received the prestigious Bell Labs Prize in 2014 for his research contributions.

–By Catherine Zandonella

Download PDF