Radhika Nagpal

Assistant Professor of Computer Science
School of Engineering and Applied Sciences
Harvard University
235 Maxwell Dworkin
33 Oxford Street,Cambridge, MA 02138
Phone: 617-496-6434
rad at eecs harvard

Research Interests: Engineering and understanding self-organizing systems. Biologically-inspired multi-agent algorithms, and their application to distributed systems, modular/swarm robotics, sensor networks, and programmable materials; Computational models of biological multicellular systems in development and morphogenesis.

Teaching (Fall 2008): CS 266:Bio-inspired Distributed and Multi-agent Systems

Short bio: In fall 2004, I joined Harvard, as an assistant professor in Computer Science, in the Harvard School of Engineering and Applied Sciences. Before that I spent a year as a research fellow in the Department of Systems Biology at Harvard Medical School. Previously, I was a postdoctoral lecturer and graduate student at the MIT Computer Science and Artificial Intelligence Laboratory (CSAIL) and a member of the Amorphous Computing Group. I am also a recipient of the 2005 Microsoft New Faculty Fellowship Award and the NSF Career Award 2007.

Contact: Office hours TUES 3-4:45pm (starting 2/19). If you are a Harvard Student, then I strongly recommend coming by during Office Hours. If you are a prospective student, intern or postdoc, please read this first. For all else, email is the best way to contact me.

My Research   |   Official Research Group page   |   Publications   |   Teaching   |   Undergrad Clubs (RoboCup, iGEM)   |   Other Fun Stuff


News


Chih-han and Radhika's paper at AAMAS 2008 is nominated for the Best Student Paper Award.

The Self-Adapting Modular Robot (and Radhika) get featured in the Faculty Profile in SEAS Newsletter and the Microsoft New Faculty Fellowship Video.

Radhika gets the NSF CAREER Award (SEAS homepage, Gazette Article), 2007.

Emergent geometric order in epithelial tissues, Nature, Aug 31, 2006.

Soren Lund, Director of Lego Mindstorms, visits the SSR Robotics lab, April 2007.

Watch the video of Harvard-MIT 2007 Robot soccer team (aka RFC Cambridge) on youtube.

Self-Organising Systems Research

Biological systems get tremendous mileage by using vast numbers of cheap and unreliable components to achieve complex goals reliably. For example, cells with identical DNA cooperate to form complex structures, such as ourselves, with incredible reliability in the face of cell death, variations in cell numbers, and changes in the environment. Emerging technologies have made it possible to create our own large-scale multi-agent systems, by bulk-manufacturing tiny computing and sensing agents and embedding these into materials and the environment. We would like to build novel applications from these technologies --- vast sensor-rich environments, robot swarms and reconfigurable modular robots, programmable materials. A key challenge is achieving the kind ofreliability and complexity that cells achieve:

  • How does one creat globally robust systems, from the cooperation of vast numbers of unreliable agents?
  • How does one translate desired global goals into the local interactions of vast numbers of agents?

My research interest is developing programming paradigms for robust collective behavior, inspired by biology. Developmental biology, how cells cooperate in tissues and multicellular organisms, can provide insights into how global self-repair and adaptation can be achieved through simple local behaviors. The study of social insects can teach us how to program cooperation and adaptation amongst mobile agents. Ultimately, the goal is to create a framework for the design and analysis of self-organising multi-agent systems. My group's approach is to formalize these strategies as algorithms, analysis, theoretical models, and programming languages. We are especially interested in global-to-local compilation, the ability to specify user goals at the high level and automatically derive provable strategies at the agent level. This methodology is applicable to a wide range of distributed multi-agent systems, from wireless sensor networks to modular and swarm robotics, and we pursue both theory and physical implementations of our work, especially in robotics.

My other research interest is understanding robust collective behavior in biological systems. Building artificial systems can give us insights into how complex global properties can arise from identically-programmed parts --- for example, how cells can form scale-independent patterns, how large morphological variations can arise from small genetic changes, and how complex cascades of decisions can tolerate variations in timing. I am interested in mathematical and computational models of multi-cellular behavior, that capture hypotheses of cell behavior and cell-cell interactions as multi-agent systems, and can be used to provide insights into systems level behavior that should emerge. We work in close collaboration with biologists, and currently study growth and pattern formation in the fruit fly wing.

Selected Publications:

  • Programmable Self-Assembly Using Biologically-Inspired Multiagent Control, AAMAS 2002. (pdf)
  • Self-repair and Scale Independent Self-reconfiguration (for a Modular Robot), IROS 2004.(pdf)
  • Firefly-Inspired Sensor Network Synchronicity, SenSys 2005. (pdf)
  • Distributed Construction by Mobile Robots with Enhanced Building Blocks, ICRA 2006. (pdf)
  • The Emergence of Geometric Order in Proliferating Metazoan Epithelia, Nature, Aug 31, 2006.

Media:

Publications,  CV,  PhD Dissertation

For more detailed description of the projects we work on see the Research Group Webpage, 



Teaching:
CS 51:Introduction to Computer Science II
(Spring 2005-08) This course is about Abstraction and Design; understanding how to use abstractions to design programs that are clear, efficient and elegant. We cover abstract models for computational processes and their concrete realization --- functional abstraction, data abstraction, object-oriented design, and finally programming languages as the ultimate abstraction.
CS 266:Bio-inspired Multi-agent and Distributed Systems
(Fall 04-08) This class will survey the state of the art in biologically-inspired approaches to designing robust collective behavior, in diverse domains. Topics include: swarm intelligence and applications, amorphous computing and reconfigurable robotics, immune-inspired systems, synthetic biology. Students will lead discussions of research papers and undertake a semester-long research project.
SB 301: Special Topics in Systems Biology
An exploration of new directions for the field of systems biology. We discuss unsolved questions in biology and new approaches offered by systems biology. Topics included theory of biological networks, understanding multicellular systems, genomics as a toolkit, and non-genetic variation. Instructors: Galit Lahav, Kit Parker, Radhika Nagpal, Vamsi Mootha, Andrew Murray, Carl Pabo.
Courses taught at MIT, as a postdoctoral lecturer
MIT 6.042: Mathematics for Computer Scientists: Fall01-02, with Prof. Albert Meyer. ( MIT OpenCourseWare version)
MIT 6.978: Biologically Motivated Programming Technology for Robust Systems: Fall 2002. (final projects)
MIT 6.033: Computer Systems Engineering: Spring 2003, Recitation Instructor.
MIT 2004 IAP Course on Synthetic Biology, with Drew Endy, Tom Knight, and Pam Silver.


Undergraduate Clubs/Activities

Harvard College Engineering Society (HCES)
RFC Cambridge, the Harvard-MIT undergraduate RoboCup soccer team, took their team of 5 autonomous robots to play soccer in the Robocup US Open held at Georgia Tech (apr 22-23). They placed 2nd and scored 3 goals!! Then they headed to Germany to compete in the International Robocup Competition!. Our very own RFC Members were even featured on CARTOON NETWORK! See some actual event movies on the RFC website. The Harvard College Engineering Society (HCES) is the umbrella undergraduate group aimed at promoting engineering and cross disciplinary collaboration on campus - i.e. We do COOL things. Come join us!
Faculty Advisors: Howard Stone, Radhika Nagpal, Robert Wood




iGEM 2006: Intercollegiate Genetically Engineered Machine Competition
Each year many universities compete in iGEM --- to design, build and characterize genetically-encoded finite state machines. The goal is to create 'living' machines by instructing cells to for example count, decode signals, or produce specific patterns. The challenge is to go from idea, to design, to DNA, to implementation (in cells) in 3 months! The Harvard Team consists of undergraduates from many different disciplines, and they have built many things from bacteria that can propogate a pulse, to dna structures that self-assemble. Read about it here: Harvard 2006 team, Harvard 2005 team and in the Gazette (Aug 25, 2005) and The New York Times. The Harvard iGEM 2006 team and one of their cool projects on the construction of novel DNA nanostructures for drug delivery made the front page of MIT's Technology Review's website.
Faculty Advisors: Pam Silver, George Church, William Shih, Radhika Nagpal, Jagesh Shah, Kit Parker, and many others


Other Fun Stuff