Prospective Members

In the modern world, autonomous control systems keep the lights on and the planes in the air. Soon, they will also be keeping cars on the road and robotic explorers operating in space and on other planets. And yet, as autonomous systems become at once more common and more complex, what guarantees do we have about their safety and reliability?

In our lab, we look at autonomy through two complementary points of view:

  1. Given an autonomous system, how do we know whether it will accomplish their missions safely?
  2. How can we use safety considerations to design better autonomous systems?

Currently, we are thinking a lot about how these questions can apply to learning-enabled autonomy — how can we check the safety of deep-learning-based controllers, and how can we train safer controllers — but we are also interested in exploring these questions more broadly.

Here are some examples of research projects from our lab:

Using neural networks to find safety certificates

Using neural networks to find safety certificates for satellite collision avoidance [arxiv](https://arxiv.org/abs/2109.06697)

Developing new reinforcement learning algorithms

Developing new reinforcement learning algorithms to safely optimize an electric vehicle transportation network [arxiv](https://arxiv.org/abs/2106.12764)

Learning safe distributed control for a 1000+ drones delivery

Learning safe distributed control for a 1000+ agent drone delivery system [arxiv](https://arxiv.org/abs/2101.05436)

Explanable AI for realistic traffic user simulation

Using explanable AI to improve the realism of traffic user simulation models [arxiv](https://arxiv.org/abs/2109.06689)

If you are interested in joining us, either as a student or postdoc at MIT, we’d love to hear from you! To help us find a fit between your interests and our work in REALM, please follow the instructions below.

Prospective Postdoctoral Researcher

Postdoc candidates are encouraged to reach out to Prof. Fan (chuchu at mit dot edu) with the following information:

  • Your CV and a list of publications.
  • A description of your research interests, and how you see those aligning with our work in REALM.
  • Names and contact information for two academic references.

Prospective PhD Students

Students interested in working with REALM should apply to the Department of Aeronautics and Astronautics at MIT and indicate an interest in autonomy and our lab in particular on your application. We can also support students applying to related departments (e.g., Mechanical Engineering and EECS), but AeroAstro is our home department. We are not able to offer specific advice on your application materials, but you may be able to take advantage of the AeroAstro Graduate Application Assistance Program (GAAP; see here for more information). If you have questions about our research, you may contact Prof. Fan (chuchu at mit dot edu) and include the following:

  • “REALM inquiry (prospective graduate student)” somewhere in the subject line.
  • A few sentences about why you’re interested in our work, and any projects that caught your attention.
  • Your CV, undergraduate transcript, and graduate transcript.

Prospective students are encouraged to apply for fellowships; please see this list of fellowship resources from MIT.

Current MIT Undergraduates

We regularly take students through MIT’s UROP program; if your are interested in pursuing a UROP in our lab, please email Oswin So (oswinso at mit dot edu). Please include the following:

  • “REALM inquiry (UROP)” somewhere in the subject line.
  • A few sentences about why you’re interested in our work, and any projects that caught your attention.
  • A few sentences about your goals for a research experience (e.g. prepare for graduate school, gain robotics experience, learn about learning-based controls, etc.).
  • A few bullet points detailing your relevant experience (e.g. robotics or controls classes, coding experience). When considering UROP students, passion counts just as much as experience, so don’t worry if you don’t have any specifically relevant experience yet.

Current MIT MS/PhD Students

Prof. Fan teaches Course 16.332 during the spring semester focusing on formal methods for autonomy. If you are interested in learning more about our work, this class is a good starting point. You may also contact Prof. Fan (chuchu at mit dot edu) or any of the graduate students in our lab to learn more about their work. We welcome collaborations!