Complex Systems, Network Science and Computation
Solving the worlds most challenging problems.
As far as anthropologists can tell with much certainty, hominids have been making tools for at least 2.6 million years. Those earliest stone tools were not much more than fist-sized rocks used to pound whatever needed to be pounded and flakes of quartz, obsidian and flint whose sharp edges were helpful in cutting whatever needed to be cut.
It is now clear humans are not alone in our ability to shape and use objects to make some tasks easier. If you watch video of a sea otter using a stone to crack open clams or an elephant using a log to get at some food hung out of reach or a chimp using a twig to fish termites out of mounds you can understand immediately how the tools work. Everything about the technology is right there on the surface for all to see and understand.
But what about the computer, smart phone or tablet device you are using right now to read these words? Can the average person look at the technology and immediately understand how it works? Sure, people can speak in general terms about the basics of binary code, silicon chips and wireless signals, but very few people have any real understanding of how any of it works. For most of our 2.6-million-year, tool-using, hominid history, most members of the species could understand even the most complex technologies in use. With the industrial revolution, that came to a screeching halt.
The tools industrialized society relies on today most heavily are, more often than not, technologies that result from, and rely on, complex systems and networks of many interacting parts. Those parts can be the arrays of satellites and antennas that enable our travel, communications, weather forecasting and earthquake research. They can be the farmers, transportation networks, safety inspectors and outlets that distribute food. They can be the miners, drillers, pipelines, power plants, wind farms, solar arrays and power grid that supply electricity. Just about every aspect of modern life in the industrialized world is supported by a complex network of people and technologies.
Cornell Engineering, with its deep pool of experts in all engineering fields, nationally-recognized research centers and advanced technology, is uniquely suited to find answers to the pressing questions and challenging problems raised by these complex systems.
Complex systems—whether integrated circuits, information relays, transportation routing, social systems, or biochemical reactions in a living cell—all behave in ways that cannot be fully explained by understanding their component parts. Complex systems require integrative approaches. Tools are needed to characterize and design these nonlinear systems that are more than the sum of the properties of their parts. The wide variety of complex systems known as networks require analytic methods of discovery by computer scientists, applied mathematicians, and engineering faculty from across the college.
The Earth Itself is a Complex System
Some of the researchers at Cornell Engineering take the Earth itself as the complex system they study. David Hysell, professor of Earth and atmospheric science (EAS), looks at the ionosphere and its effect on human activities and communications.
“The ionosphere is a spectacularly complicated medium,” says Hysell. “It looks different from every angle. Scientifically, it is great because it is so complicated and so rewarding. I get to feed my intellectual side because I am exploring some really basic scientific questions and at the same time my work has immediate practical applications.”
Much of the early interest in the ionosphere was driven by NASA and the U.S. Department of Defense. They were flying their vehicles through an environment they did not know much about. And then, more and more of the systems humans use to communicate began to pass through the ionosphere, and understanding its properties and behavior became even more essential. The ionosphere has its own ‘weather,’ but it is not the same sort of weather we experience at the Earth’s surface. The climate and weather in the ionosphere are fluctuations in energy and electron concentrations driven by Earth’s response to emissions from the sun.
The stunning increase in computing power over the past few years has enabled Hysell and other researchers to use the data collected at the Jicamarca Radio Observatory in Peru to begin to model the ionosphere. “This system is so complex that today’s computers are still inadequate,” says Hysell, “but they do now allow us to keep track of a billion particles through code. We are starting to be able to measure state parameters and we are beginning to get a grip on how the most important variables behave normally.”
Hysell believes the day is coming soon when we will be able to produce accurate forecasts of ionospheric weather. “Using the data from Jicamarca and comparing our predictions with observed reality, we are slowly getting to a good model of what goes on in the ionosphere,” says Hysell. “Our goal is to safeguard against outages of services people rely on, and that day is getting closer.”
Other EAS researchers are using the tools of seismicity and geodesy to study the complex motions of the Earth under our feet. Associate Professor Rowena Lohman hopes to build a greater understanding of how all the things people do at or below the surface impact seismicity. “Sometimes surface deformation precedes earthquakes, particularly in volcanic regions” says Lohman, “but usually it doesn’t. Understanding the difference is very important.” Excavating mines, tapping into aquifers, building dams to create reservoirs, hydrofracking, CO2 sequestration, deforestation and urban development all deform the surface—some very slowly and others fairly quickly. Collecting data on these deformations and analyzing the data as it relates to earthquakes can give researchers a better idea of how human activities affect seismicity. This idea has been in the news a lot lately due to swarms of earthquakes in the normally seismically inactive bedrock of Oklahoma.
With her research, Lohman hopes to increase what we know about the planet we live on. “Cornell’s Department of Earth and Atmospheric Sciences is so strong,” says Lohman. “It is a great place to be. We have people working in the area of geodesy and others focused on seismicity and because it is such a collaborative place, we can bring these things together in a way to that will greatly increase what we know about the Earth.”
The Tangled Webs We Weave
Researchers from the School of Civil and Environmental Engineering (CEE) focus some of their work on the intersection of the Earth and the complex systems humans build there. Professor Emeritus Peter Loucks has spent a career applying economic, ecological and systems analytical methods to the solution of environmental and water resource problems around the world.
Loucks has created tools for water resource management that take into account the many complicated factors beyond simply the amount of water available in a region. He has been so successful in helping businesses, municipalities, and regional and national governments address water issues that he was recently awarded the Prince Sultan Bin Abdulaziz International Prize for Water. The official announcement praises Loucks for creating “an effective, dynamic framework which is used successfully throughout the world to examine the interplay between environmental stress, stakeholder participation processes and hydrological systems.” It goes on to say, “Decision makers in numerous countries, including developing nations, have been trained and influenced by Dr. Loucks’ approach to water resources planning. This is precisely why he is known as the ‘father of the systems approach to water resources management.’”
A Cornell researcher at the other end of the career spectrum who also takes a systems approach is new CEE Assistant Professor Samitha Samaranayake. Samaranayake believes that we’re at the dawn of a revolution in urban transportation systems. “This is a great time to study urban transportation problems and solutions,” says Samaranayake. “Bike-share and car-share programs are springing up everywhere. Cities are innovating and trying to be more agile with public transportation. On-demand systems like Uber and Lyft are becoming popular. Many people in urban environments are rethinking the need for personal vehicle ownership. Combine all these with the data we can collect in real time from smart phones and internet-enabled devices and you have a real opportunity to create some great solutions.”
Samaranayake is interested in the modeling, analysis and control of networked urban infrastructure systems with a focus on transportation networks. His long term goal is to help large urban areas create truly integrated multi-modal transit systems that move people efficiently and reduce energy usage and congestion. “One of the hardest problems, at the end of the day, is understanding and modeling human behavior,” says Samaranayake. Therefore, effective solutions will require collaborations not only with engineers, but also with behavioral economists, public policy experts and others from the social sciences. He decided to come to Cornell, in part, because of the well-founded reputation Cornell has as a place where researchers from many fields collaborate willingly and profitably.
Tom O’Rourke, the Thomas R. Briggs Professor of Engineering, has focused his work on protecting the underground built systems humans use to transport water, gas, electricity and themselves. O’Rourke’s work runs the gamut from developing analytical methods and siting strategies to mitigate pipeline and tunnel damage during earthquakes, to the development and application of advanced polymer and composite materials for the in-situ rehabilitation of water supply and gas distribution systems. Like Loucks, O’Rourke has been recognized for his impressive lifetime achievements.
“We operate the Cornell Large-Scale Lifelines Testing Facility” says O’Rourke. “This is a world class, unique facility that is able to simulate fault rupture effects on underground infrastructure. We are working with the water, natural gas and pipeline industries to develop the next generation hazard-resilient underground pipelines and in-situ rehabilitation systems. We are totally supported by industry and qualify about one new product every six months that goes directly into water delivery systems on the West Coast to improve earthquake resilience.”
O’Rourke and his colleagues are developing the next generation soil-structure interaction analytical models so that designers and planners have the computational tools to implement the new hazard resilient pipelines. Not only does Cornell have the capability of testing to confirm performance at the actual scale of pipelines installed in the field, but also is able to simulate the response of real-world pipeline systems. For example, O’Rourke and his co-workers have simulated all 12,000 kilometers of distribution pipelines and related facilities—tanks, reservoirs, pressure regulation stations—in the Los Angeles Department of Water and Power system to see how pipeline and operational improvements increase the reliability of the system under earthquake effects.
In April 2016 O’Rourke was awarded the George W. Housner Medal by the Earthquake Engineering Research Institute. The Housner Medal recognizes those who have made extraordinary and lasting contributions to public earthquake safety through the development and application of earthquake hazard reduction practices and policies. It is the most prestigious award of the institute. In his 40 years at Cornell, O’Rourke has authored or co-authored over 370 technical publications. His research interests cover geotechnical engineering, earthquake engineering, underground construction technologies, engineering for large, geographically distributed systems, and geographic information technologies and database management. His body of work gives full recognition to the complex nature of the systems humans have created to enable the modern technological lives we live.
While some of the complex systems Cornell Engineering researchers focus on are as grand as the planet itself, others might someday be invited into your home to act as a personal assistant. If you think about the many interdependent systems necessary to create safe, reliable and useful robots, you can see why Cornell would be an excellent place for robotics research. It takes experts in mechanical engineering, materials science, electrical and computer engineering, computer science and other fields as far-flung as linguistics and sociology to begin to create robots that can become truly useful in everyday life.
Within the robotics group at Cornell Engineering there are people studying estimation theory and control for autonomous and semi-autonomous systems, computational intelligence and sensometer learning, human-robot teamwork and collaboration, advanced factory automation, verifiabilitycorrect robot behavior, multi-robot autonomous collectives capable of construction, energy-efficient robotic motion, and the development of soft actuators and machines. In other words, everything from parts to programs.
Two new faculty members at Cornell are doing work right at the frontier of robotics. Assistant Professor and Mills Family Faculty Fellow Guy Hoffman in the Sibley School of Mechanical and Aerospace Engineering sees a future where humans and robots will find themselves working together on tasks that require cooperation. He has been working with psychologists, among others, to create a behavioral model of what humans would look for in a good robotic companion. “Surprisingly, people are not always happy with a robotic teammate who is entirely rigid and predictable,” says Hoffman. “People value flexibility, even in a robot. It can be delightful to be surprised.” One major question Hoffman will need to answer is: how does a robot balance initiative and flexibility with reliability?
Kirstin Petersen, assistant professor in the School of Electrical and Computer Engineering, takes her inspiration from the naturally-occurring complex systems certain insect species create. Petersen is interested in the design and coordination of bio-inspired robot collectives. The task assigned to her initial robot swarms is construction. Petersen draws her inspiration from ants, bees, termites and other social insects. In these species, thousands of individuals coordinate to gather food, construct nests and defend the group. No individual insect could accomplish any of these tasks alone, yet the group manages to do all of this without central control.
Petersen is designing her robot swarm to mimic the scalability and error tolerance of insect systems. Another goal of Petersen’s is to replace the need for complicated sensors and control with passive mechanical features. “These robot swarms manage to combine many of my research interests,” says Petersen. “To build them requires knowledge of mechanics, electronics, computer science, biology and more. They are a great challenge.”
Better Chemistry Through Modeling
In the field of chemical engineering, researchers are applying the ideas of systems design and mathematical modeling to push the field into unexplored areas.
Fengqi You is the Roxanne E. and Michael J. Zak Professor and a David Croll Faculty Fellow in the Robert Frederick Smith School of Chemical and Biomolecular Engineering. “I believe that chemical engineers should be able to do everything,” says You. “Society faces many issues in the areas of food, water and energy production. Chemical engineers can help come up with sustainable solutions to important challenges in these areas.”
You’s focus is on the development of novel computational models, optimization algorithms, and systems analysis and design methods to improve process manufacturing, energy systems and sustainability. His work straddles the line between operations research and chemical engineering. “Michael Zak ’75 endowed the faculty chair I am in because he believes strongly that it is essential to link the field of operations research with chemical engineering,” says You.
His long-term goal is to create computational and modeling tools that decision-makers in government and industry will be able to use to make better decisions with the idea of sustainability as a guiding factor. “We have known for a long time that people need to make better decisions about where and how we get our food, water and energy, yet our decisions are not as good as we expected” says You. “I want to develop tools that will be analogous to the tools weather forecasters use today.” You believes that if he can use advanced mathematical, computing and optimization techniques to create systems engineering tools, then he can provide decision-makers with the tools they will need to make better decisions.
You believes that now is the perfect time to create these tools. “With more and more data available to researchers it is time to take that data and push the limits of sustainability analytics,” says You. Processes used to manufacture biofuels, solar cells, shale gas and many other products could be made more efficient and more sustainable. With his knowledge of chemical processes and his background in operations research, You is ideally situated to take advantage of the big data now available to help the world better manage its resources while at the same time helping companies save money through more efficient operations.
You’s colleague in the Smith School, Professor Jeffrey Varner, writes computer code that models the flow of signals and materials in biological signal transduction and metabolic networks. When a molecule from outside a cell activates a receptor on the cell surface, a chain of biochemical events is triggered, leading to a cellular response. Even though cells are the basic unit of life, their biochemistry is anything but basic. Modeling the chain of reactions that takes place as a signal makes its way from the surface to the nucleus of a cell is a complex undertaking. Once the signal reaches the nucleus, the cell may change its metabolism, change its shape, alter the expression of certain genes or change in its rate of division. Understanding the flow of these signals through cells could help researchers learn how to reprogram it. In effect, we could start to tell the cell what to do.
Of particular interest to Varner has been the biochemistry of human blood. He is working to build a better understanding of the processes that occur following injury; for example blood coagulation, or infection. His eventual goal is to be able to use what he learns to create effective treatments to control hemorrhage. “Few people know that trauma is the leading cause of death for persons 36 years old and younger and that hemorrhage accounts for 40 percent of all trauma deaths,” says Varner. “I want to understand how the body responds to trauma and how we can manage this response to save lives.”
“I love the biology behind all these processes,” says Varner. “But what I really bring to the table is the ability to explain the biology through mathematical modeling.”
A Smarter Grid and a Cleaner Energy Future
One of the most ubiquitous and essential networks in the industrialized world is the power grid. Due to the broad-based strength of Cornell Engineering, the university has an international reputation in power systems research. In 1996 Cornell was one of the founding members of the Power System Engineering Research Center (PSERC)—a consortium of industry, government and university members working together to address the basic engineering and economic issues facing the rapidly evolving electric grid.
Cornell’s contingent in the PSERC includes researchers from the Department of Biological and Environmental Engineering, the Sibley School, the Charles H. Dyson School of Applied Economics and Management, the Department of Computer Science, and the School of Electrical and Computer Engineering. PSERC research has multiple directions, including improving grid operational efficiency, security and reliability; real-time power system assessment, control and optimization; large-scale integration of variable renewable energy resources; energy economics and market design; power system stability and contingency analysis; cyber-security and privacy; demand response and energy management; decentralized energy management; and microgrids.
Improving the grid is just one piece of the work Cornell Engineering is doing in the essential area of energy research. The Cornell Energy Institute takes the lead at Cornell Engineering in technology-based research and education in energy. Reflecting the interconnected nature of the energy challenges facing the world today, the Energy Institute collaborates across Cornell with faculty from social, physical and natural sciences.
The institute is led by CBE Professor Jefferson Tester, who holds the Croll Professorship of Sustainable Energy Systems. Researchers in Cornell’s Energy Institute study all aspects of the production, transfer and storage of energy. Their work ranges from the design and production of new materials for solar cells and batteries to the mathematical modeling required to create the optimal arrangement of turbines in massive offshore wind farms. It is clear that humans need to find new and better ways to create, use, transfer and store power. For example, CBE Professor Fengqi You’s research into the life cycle aspects of manufacturing solar PV modules and converting biomass feedstocks to biofuels is providing a means for making informed decisions regarding specific technology options.
To achieve the maximum benefits of using renewable energy to transform an energy system, careful integration is necessary. It will be essential to match demands for heating, electricity and transportation fuels with regionally available renewable energy resources and careful consideration of environmental and economic tradeoffs. The undertaking is monumental, as are the consequences of failure. The complexity of the systems involved is daunting, but everywhere you look at Cornell, you find people qualified to examine that complexity, make sense of it, and create solutions.
Big Messy Data
One hallmark of complex systems and networks is the amount of information both required to operate the system and available to provide insights for how to operate the system better. Often called “big data,” researchers, coders and systems engineers at Cornell just as often make a point of calling it “big messy data.” Madeleine Udell, assistant professor and Richard and Sybil Smith Sesquicentennial Fellow at Cornell’s School of Operations Research and Information Engineering (ORIE), is even teaching a class called Learning With Big Messy Data (ORIE 4741).
Udell’s research at Cornell focuses on discovering structure in large messy data sets, and designing efficient optimization schemes for novel applications in data analysis and engineering system design. Put more plainly, Udell creates modeling software that can help people in finance and medicine use the data they collect to make improvements to how they do their work. “I question people in industry to better understand the problems they need to solve,” says Udell. “I look for similarities across industries so instead of addressing one problem at a time I can solve a whole category of problems. I find ways of using math to make better decisions. No matter who I talk to in any field, it turns out they are solving some optimization problem.”
If you are unfamiliar with the idea of optimization, you should get ready to hear the term a lot more in the coming years. With the massive increase in the amount of data collected, combined with the improved speed and power of computation, optimization is quickly becoming an essential tool. To optimize a process is to find the most cost effective or highest achievable performance under given constraints.
Researchers who work in operations research, information engineering, computer science and computation create models and algorithms that can help determine the best allocation of bikes throughout the city for New York’s Citibike program and the best allocation of hedge fund capital under specific economic conditions. They determine where to station ambulances around a city for best response time, or even which chemical compounds might make the most likely ingredients in future drug therapies.
One researcher whose work exemplifies the power of data to help companies, governments, and scientists manage complex real-world systems with maximum efficiency is ORIE Associate Professor Peter Frazier. Frazier has worked with Yelp to build a software platform for improving websites and apps, which is also used by Netflix. He has worked with Uber to help make their operations better. He has worked with IBM and Weill Cornell Medicine to design personalized medical testing procedures for patients with vascular disease. He has worked with other companies, like 3M, Jane Street Capital, and an Ithaca company called Nature Source Genetics to address other difficult problems those companies face.
One line of research Frazier is most excited about is his collaboration with chemists working on nanomaterials and drug discovery. Drug and materials experiments can be very expensive to run and the results are in no way guaranteed. Frazier uses statistics and optimization to help scientists design experiments that will maximize the value of the information they can collect. Even though his Ph.D. is in operations research, his background in physics has made it easier for Frazier to communicate across fields with biochemists, medicinal chemists and materials scientists.
This, together with the power of his operations research methods, makes him a valuable collaborator in biochemical research. “Based on mathematical models, I give guidance on which particular molecules might work best in a given experiment,” says Frazier. “Also, I can help determine which experiment from a list of several possibilities would be the best to run.”
The Computer Age
And now we come back around to the device you are using to access this article. Whether it is a smart phone, a tablet, a laptop or a desktop computer, it is a marvel of engineering that is the physical manifestation of some incredibly complex systems and networks. Through the World Wide Web, you have direct access to more information than anyone at any other time in human history. You can use this device to communicate in real time with people all around the world. You can pay bills and buy items that could appear outside your door tomorrow. Of course, you could not do much with this device without the code that tells it what to do and how to do it.
From the School of Electrical and Computer Engineering, Professor Stephen Wicker focuses on the interface between information networking technology, law and sociology. His research shines a light on how design choices and regulatory frameworks can infringe the privacy and speech rights of users. “Given our use of cellular technology in every facet of our lives, security and privacy have never been more important,” says Wicker.
Researchers at Cornell’s Department of Computer Science have been making computers and computer networks better, faster and safer since the department was founded in 1965. It is one of the oldest computer science departments in the world and right off the bat, faculty were doing foundational work. Gerard Salton’s work on information retrieval laid the groundwork for Google and other search engines of today. Department Chair Juris Hartmanis published the paper that started the field of computational complexity in 1965.
Cornell’s computer scientists today have continued the rich traditions begun by Salton, Hartmanis and others more than 50 years ago. Emin Gün Sirer, associate professor at Cornell, is interested in making cryptocurrencies safer, better understood and more commonly accepted by individuals, businesses and governments. His research has laid the groundwork for currencies such as Bitcoin, as well as other consensus systems that connect businesses at very large scales.
Department Chair and Samuel B. Eckert Professor Fred Schneider’s recent work focuses on system security. In addition to teaching and research, he frequently consults with industry. In this way he gets firsthand knowledge of the problems and challenges facing real world companies. He provides technical expertise in fault-tolerance and computer security to the companies he consults with. He has also testified about cybersecurity research at hearings of the U.S. House of Representatives Armed Services Committee, as well as the Committee on Science and Technology.
Complexity is Everywhere
The examples discussed above merely begin to paint a picture of the work done at Cornell Engineering in the broad area called complex systems, networks and computation.
In addition to Cornell Engineering’s Ithaca campus, several Cornell programs are located in New York City, where unique opportunities exist to solve complex computational problems. ORIE and the Department of Information Science have teamed up with Cornell Tech to prepare students to create and use cutting-edge algorithms and computation to transform large, unstructured data sets into business insights for the city’s rapidly growing tech sector. The school also operates Cornell Financial Engineering Manhattan. This satellite campus in New York’s financial district focuses on machine learning applications for hedge fund management, stock trading and bank regulation.
Understanding just how complex many of today’s essential technologies and systems really are can make a person shake their head and say with a bemused smile, “where would you even start?” That understanding of the utter complexity of the modern world makes Cornell Engineering researchers respond with those very same words, and an eager smile: “Yeah, where would you start?”