News

"Neural Lander" Uses AI to Land Drones Smoothly

05-23-19

Professors Chung, Anandkumar, and Yue have teamed up to develop a system that uses a deep neural network to help autonomous drones "learn" how to land more safely and quickly, while gobbling up less power. The system they have created, dubbed the "Neural Lander," is a learning-based controller that tracks the position and speed of the drone, and modifies its landing trajectory and rotor speed accordingly to achieve the smoothest possible landing. The new system could prove crucial to projects currently under development at CAST, including an autonomous medical transport that could land in difficult-to-reach locations (such as a gridlocked traffic). "The importance of being able to land swiftly and smoothly when transporting an injured individual cannot be overstated," says Professor Gharib who is the director of CAST; and one of the lead researchers of the air ambulance project. [Caltech story]

Tags: research highlights Morteza Gharib Yisong Yue Soon-Jo Chung Animashree Anandkumar

Katie Bouman Joins EAS and CMS

04-11-19

Congratulations to the entire Event Horizon Telescope team, and especially to Dr. Katie Bouman who is joining the Engineering and Applied Science (EAS) Division in June as assistant professor of computing and mathematical sciences (CMS). Currently, Caltech and CO Architects are working with her to design and construct a unique laboratory that will facilitate her work in computational imaging. The laboratory is the first of its kind and is designed for her to conduct experimental work in conjunction with her computational approaches – making it possible, for instance, to observe phenomena previously difficult or impossible to measure. The black hole imaging is one spectacular example of how Professor Bouman’s algorithms are advancing our knowledge of the world; she has also developed algorithms that let us “see around corners” and detect material properties (such as stiffness and dampness) via imaging. In her work, Bouman has also developed methods to combine information from both imaging as well as acoustic systems to analyze sub-pixel scale vibrations of otherwise seemingly still objects. As a result, relatively inexpensive cameras, combined with powerful algorithms, are an increasingly attractive alternative to complex and expensive laser-based systems to sense “invisible” attributes of a material. [Caltech story - How to Take a Picture of a Black Hole]

Tags: research highlights CMS Katie Bouman

Computer Scientists Create Reprogrammable Molecular Computing System

03-20-19

Erik Winfree, Professor of Computer Science, Computation and Neural Systems, and Bioengineering, and colleagues have designed DNA molecules that can carry out reprogrammable computations, for the first time creating so-called algorithmic self-assembly in which the same "hardware" can be configured to run different "software." Although DNA computers have the potential to perform more complex computations than the ones featured in the Nature paper, Professor Winfree cautions that one should not expect them to start replacing the standard silicon microchip computers. That is not the point of this research. "These are rudimentary computations, but they have the power to teach us more about how simple molecular processes like self-assembly can encode information and carry out algorithms. Biology is proof that chemistry is inherently information-based and can store information that can direct algorithmic behavior at the molecular level," he says. [Caltech story]

Tags: research highlights CMS Erik Winfree

Creating a "Virtual Seismologist"

01-04-19

Professor Yisong Yue is collaborating with Caltech seismologists to use artificial intelligence (AI) to improve the automated processes that identify earthquake waves and assess the strength, speed, and direction of shaking in real time. Professor Yue explains, “the reasons why AI can be a good tool have to do with scale and complexity coupled with an abundant amount of data. Earthquake monitoring systems generate massive data sets that need to be processed in order to provide useful information to scientists. AI can do that faster and more accurately than humans can, and even find patterns that would otherwise escape the human eye.” [Read the full Q&A]

Tags: research highlights CMS Yisong Yue Egill Hauksson Zachary Ross Men-Andrin Meier

New Climate Model to Be Built from the Ground Up

12-13-18

"Projections with current climate models—for example, of how features such as rainfall extremes will change—still have large uncertainties, and the uncertainties are poorly quantified," says Professor Tapio Schneider, principal investigator of the Climate Modeling Alliance (CliMA). "For cities planning their stormwater management infrastructure to withstand the next 100 years' worth of floods, this is a serious issue; concrete answers about the likely range of climate outcomes are key for planning." The new climate model will be built by a consortium of researchers led by Caltech, in partnership with MIT; the Naval Postgraduate School (NPS); and JPL, which Caltech manages for NASA. It will use data-assimilation and machine-learning tools to improve itself in real time, harnessing both Earth observations and the nested high-resolution simulations. "The success of computational weather forecasting demonstrates the power of using data to improve the accuracy of computer models; we aim to bring the same successes to climate prediction," says Professor Andrew Stuart. [Caltech story]

Tags: research highlights CMS ESE Tapio Schneider Andrew Stuart

The Big Picture

12-03-18

Thanks to Professor Pietro Perona and his graduate students including Grant Van Horn and Sara Beery, the next wildlife photo you snap might set you on a path to helping map life on Earth. “The whole web, this huge repository of wonderful information, is indexed by words,” Perona says. “But when we have an image—a visual query—we don’t know what to do unless there is an expert next to us. We’ve gotten so numb to the idea that we’ll never find the answer out.” [Breakthrough story]

Tags: EE research highlights CMS Pietro Perona Grant Van Horn Sara Beery

Professors Barr and Schröder Elected to ACM SIGGRAPH Academy

10-05-18

Alan Barr, Professor of Computer Science, and Peter Schröder, Shaler Arthur Hanisch Professor of Computer Science and Applied and Computational Mathematics, have been elected to the first class of the ACM SIGGRAPH Academy. Professor Barr was selected for his contributions to graphics, primarily for extending computer graphics shape modeling to include physically based and teleological modeling. Professor Schröder was recognized for his pioneering work in geometry processing and multiresolution modeling. The ACM SIGGRAPH Academy is an honorary group of individuals who have made substantial contributions to the field of computer graphics and interactive techniques. These are principal leaders of the field, whose efforts have shaped the disciplines and/or industry, and led the research and/or innovation in computer graphics and interactive techniques. [Full list of academy members]

Tags: honors research highlights CMS Peter Schröder Alan Barr

President Rosenbaum Highlights Postdocs as "Unsung Heroes"

09-24-18

In a letter to the Caltech community during National Postdoc Appreciation Week, the Caltech President emphasizes the role this key group plays at the Institute. He stated, “Caltech's mission of world-leading research and education depends crucially on our postdoctoral scholars. Although their time at Caltech may be short, they quickly become vital parts of the Institute's intellectual fabric.” [President’s Letter] [EAS Postdoc Resource Page]

Tags: APhMS EE GALCIT MedE MCE CMS ESE Thomas Rosenbaum postdocs

Caltech and Disney Engineers Collaborate on Robotics

01-18-18

Caltech and Disney Research have entered into a joint research agreement to pioneer robotic control systems and further explore artificial intelligence technologies. Pietro Perona will work with Disney roboticist Martin Buehler to create navigation and perception software that could allow robotic characters to safely move through dense crowds and interact with people. Aaron Ames will work with Disney Research's Lanny Smoot to further explore robot autonomy and machine learning by creating objects that can self-navigate and perform stunts. Yisong Yue has been working with engineers from Disney Research on the use of machine learning to analyze the behavior of soccer players and to measure audience engagement. [Caltech story]

Tags: EE research highlights MCE CMS Pietro Perona Yisong Yue Aaron Ames

Engineers Model the California Reservoir Network

11-22-17

Professor Venkat Chandrasekaran and graduate student Armeen Taeb have developed an empirical statewide model of the California reservoir network. This work offers reservoir managers insight on how to plan and respond to drought conditions. "The bread and butter of hydrology is using physical laws to describe water phenomena. But the behavior of these reservoirs is not solely determined by physical laws of the water cycle, but also by demands and what these reservoirs are being used for," Taeb explains. [Caltech story]

Tags: EE research highlights CMS Venkat Chandrasekaran Armeen Taeb