About

Chuchu Fan is an Associate Professor (pre-tenure) in the Department of Aeronautics and Astronautics (AeroAstro) and the Laboratory for Information and Decision Systems (LIDS) at MIT. Her research group, Realm at MIT, focuses on using rigorous mathematics, including formal methods, machine learning, and control theory, for the design, analysis, and verification of safe autonomous systems. She is dedicated to providing safety-critical autonomous systems with rigorous proof of their safety, efficiency, and performance. Before joining MIT, she was a postdoctoral researcher at Caltech and earned her Ph.D. at the University of Illinois at Urbana-Champaign. She completed her bachelor’s degree at Tsinghua University. Her specialization and research interests include formal methods, control, and machine learning for the design and analysis of safe autonomous systems, cyber-physical systems, and robotic systems. Her work has been recognized with several awards, including an NSF CAREER Award, an AFOSR Young Investigator Program (YIP) Award, an ONR Young Investigator Program (YIP) Award, and the 2020 ACM Doctoral Dissertation Award.

Research topics

• Computer Science
• Artificial Intelligence
• Machine Learning
• Information Retrieval
• Mathematics
• Engineering
• Computer Security
• Data Mining
• Natural Language Processing
• Theoretical computer science
• Data science
• Control engineering
• Algorithm
• World Wide Web
• Programming language

Selected publications

• Navigating the Clutter: Waypoint-Based Bi-Level Planning for Multi-Robot Systems

  arXiv (Cornell University) · 2026-04-22

  articleOpen access

  Multi-robot control in cluttered environments is a challenging problem that involves complex physical constraints, including robot-robot collisions, robot-obstacle collisions, and unreachable motions. Successful planning in such settings requires joint optimization over high-level task planning and low-level motion planning, as violations of physical constraints may arise from failures at either level. However, jointly optimizing task and motion planning is difficult due to the complex parameterization of low-level motion trajectories and the ambiguity of credit assignment across the two planning levels. In this paper, we propose a hybrid multi-robot control framework that jointly optimizes task and motion planning. To enable effective parameterization of low-level planning, we introduce waypoints, a simple yet expressive representation for motion trajectories. To address the credit assignment challenge, we adopt a curriculum-based training strategy with a modified RLVR algorithm that propagates motion feasibility feedback from the motion planner to the task planner. Experiments on BoxNet3D-OBS, a challenging multi-robot benchmark with dense obstacles and up to nine robots, show that our approach consistently improves task success over motion-agnostic and VLA-based baselines. Our code is available at https://github.com/UCSB-NLP-Chang/navigate-cluster

  PublisherOA PDF

• Navigating the Clutter: Waypoint-Based Bi-Level Planning for Multi-Robot Systems

  arXiv (Cornell University) · 2026-04-22

  preprintOpen access

  Multi-robot control in cluttered environments is a challenging problem that involves complex physical constraints, including robot-robot collisions, robot-obstacle collisions, and unreachable motions. Successful planning in such settings requires joint optimization over high-level task planning and low-level motion planning, as violations of physical constraints may arise from failures at either level. However, jointly optimizing task and motion planning is difficult due to the complex parameterization of low-level motion trajectories and the ambiguity of credit assignment across the two planning levels. In this paper, we propose a hybrid multi-robot control framework that jointly optimizes task and motion planning. To enable effective parameterization of low-level planning, we introduce waypoints, a simple yet expressive representation for motion trajectories. To address the credit assignment challenge, we adopt a curriculum-based training strategy with a modified RLVR algorithm that propagates motion feasibility feedback from the motion planner to the task planner. Experiments on BoxNet3D-OBS, a challenging multi-robot benchmark with dense obstacles and up to nine robots, show that our approach consistently improves task success over motion-agnostic and VLA-based baselines. Our code is available at https://github.com/UCSB-NLP-Chang/navigate-cluster

  PublisherDOI

• Language-Grounded Hierarchical Planning and Execution with Multi-Robot 3D Scene Graphs

  ArXiv.org · 2025-06-09

  preprintOpen access

  In this paper, we introduce a multi-robot system that integrates mapping, localization, and task and motion planning (TAMP) enabled by 3D scene graphs to execute complex instructions expressed in natural language. Our system builds a shared 3D scene graph incorporating an open-set object-based map, which is leveraged for multi-robot 3D scene graph fusion. This representation supports real-time, view-invariant relocalization (via the object-based map) and planning (via the 3D scene graph), allowing a team of robots to reason about their surroundings and execute complex tasks. Additionally, we introduce a planning approach that translates operator intent into Planning Domain Definition Language (PDDL) goals using a Large Language Model (LLM) by leveraging context from the shared 3D scene graph and robot capabilities. We provide an experimental assessment of the performance of our system on real-world tasks in large-scale, outdoor environments. A supplementary video is available at https://youtu.be/8xbGGOLfLAY.

  PublisherOA PDFDOI

• AirTaxiSim: A Simulator for Autonomous Air Taxis

  2025-07-16 · 2 citations

  article

  The rapid advancements in air mobility vehicles is paving the way for air taxis to become a viable mode of public transportation. The next technological frontier for air taxis is fully autonomous operation. Developing safe and efficient autonomous control for air taxis presents greater challenges than for ground vehicles due to the inherent instability of aerial vehicles. Therefore, simulation solutions for autonomous air taxis will play a crucial role in accelerating their development and eventual safe deployment. This paper introduces AirTaxiSim, an end to end simulation framework for autonomous air taxis. AirTaxiSim is designed to model and analyze the complexities of autonomous air taxi operations in dynamic and cluttered urban environments. AirTaxiSim integrates high fidelity physical models of vertical take-off and landing air vehicles in photo-realistic urban environments. The primary purpose of AirTaxiSim is to evaluate the safety, performance, and efficiency of autonomous air taxi services, across a variety of scenarios, including dangerous edge cases. AirTaxiSim also provides methods for generating datasets and establishing benchmarks for autonomous air taxis. This paper describes the simulator’s construction, functionalities, and some of the use cases, providing critical information to facilitate its use in advancing autonomy in aerial vehicles.

  PublisherDOI

• Collision- and Reachability-Aware Multi-Robot Control with Grounded LLM Planners

  ArXiv.org · 2025-05-26

  preprintOpen access

  Large language models (LLMs) have demonstrated strong performance in various robot control tasks. However, their deployment in real-world applications remains constrained. Even state-ofthe-art LLMs, such as GPT-o4mini, frequently produce invalid action plans that violate physical constraints, such as directing a robot to an unreachable location or causing collisions between robots. This issue primarily arises from a lack of awareness of these physical constraints during the reasoning process. To address this issue, we propose a novel framework that integrates reinforcement learning with verifiable rewards (RLVR) to incentivize knowledge of physical constraints into LLMs to induce constraints-aware reasoning during plan generation. In this approach, only valid action plans that successfully complete a control task receive positive rewards. We applied our method to two small-scale LLMs: a non-reasoning Qwen2.5-3B-Instruct and a reasoning Qwen3-4B. The experiment results demonstrate that constraint-aware small LLMs largely outperform large-scale models without constraints, grounded on both the BoxNet task and a newly developed BoxNet3D environment built using MuJoCo. This work highlights the effectiveness of grounding even small LLMs with physical constraints to enable scalable and efficient multi-robot control in complex, physically constrained environments.

  PublisherOA PDFDOI

• Analysis of convolutional neural networks and vision transformers in direct imaging for high-contrast exoplanet detection

  2025-09-16

  article

  Direct imaging of exoplanets involves collecting photons emitted or reflected by a planet orbiting a host star. This approach is challenging due to the extreme brightness difference between exoplanets and their host stars as well as the small angular separation between a star and its orbiting exoplanet. Coronagraphs and deformable mirrors are used to block and redistribute the starlight to enable imaging of the exoplanets. Speckle discrimination algorithms aim to find potential exoplanets but struggle to distinguish actual planetary signals from residual starlight speckles, which can lead to false positives. Existing speckle discrimination methods rely on binary classification rather than probabilistic outputs, limiting their ability to differentiate between exoplanets and noise based on subtle patterns. These algorithms operate in the post-processing stage and do not autonomously follow up on detected points of interest by refining observational parameters, such as spectral bands, exposure time, or region of interest. In this work, we apply deep learning models, specifically Convolutional Neural Networks (CNNs) and Vision Transformers (ViTs), to high-contrast exoplanet detection. CNNs are effective at detecting local spatial features, making them well-suited for identifying small, well-defined planetary signals. ViTs leverage self-attention mechanisms to capture long-range dependencies, which may improve their ability to distinguish exoplanets from complex noise patterns. The models are trained and tested using images from the High-contrast imager for Complex Aperture Telescopes (HiCAT) simulator developed by the Space Telescope Science Institute, with synthetic exoplanets injected into raw testbed images. By comparing the performance of CNNs and ViTs, we assess their suitability for future exoplanet detection efforts. This study highlights how AI-driven approaches can address the growing demands of next-generation observatories by enhancing detection sensitivity, reducing false positives, and enabling real-time follow-up actions to refine imaging parameters.

  PublisherDOI

• Hierarchical Multi-Agent Reinforcement Learning with Control Barrier Functions for Safety-Critical Autonomous Systems

  ArXiv.org · 2025-07-20

  preprintOpen access

  We address the problem of safe policy learning in multi-agent safety-critical autonomous systems. In such systems, it is necessary for each agent to meet the safety requirements at all times while also cooperating with other agents to accomplish the task. Toward this end, we propose a safe Hierarchical Multi-Agent Reinforcement Learning (HMARL) approach based on Control Barrier Functions (CBFs). Our proposed hierarchical approach decomposes the overall reinforcement learning problem into two levels learning joint cooperative behavior at the higher level and learning safe individual behavior at the lower or agent level conditioned on the high-level policy. Specifically, we propose a skill-based HMARL-CBF algorithm in which the higher level problem involves learning a joint policy over the skills for all the agents and the lower-level problem involves learning policies to execute the skills safely with CBFs. We validate our approach on challenging environment scenarios whereby a large number of agents have to safely navigate through conflicting road networks. Compared with existing state of the art methods, our approach significantly improves the safety achieving near perfect (within 5%) success/safety rate while also improving performance across all the environments.

  PublisherOA PDFDOI

• Solving Multi-Agent Safe Optimal Control with Distributed Epigraph Form MARL

  ArXiv.org · 2025-04-21

  preprintOpen accessSenior author

  Tasks for multi-robot systems often require the robots to collaborate and complete a team goal while maintaining safety. This problem is usually formalized as a constrained Markov decision process (CMDP), which targets minimizing a global cost and bringing the mean of constraint violation below a user-defined threshold. Inspired by real-world robotic applications, we define safety as zero constraint violation. While many safe multi-agent reinforcement learning (MARL) algorithms have been proposed to solve CMDPs, these algorithms suffer from unstable training in this setting. To tackle this, we use the epigraph form for constrained optimization to improve training stability and prove that the centralized epigraph form problem can be solved in a distributed fashion by each agent. This results in a novel centralized training distributed execution MARL algorithm named Def-MARL. Simulation experiments on 8 different tasks across 2 different simulators show that Def-MARL achieves the best overall performance, satisfies safety constraints, and maintains stable training. Real-world hardware experiments on Crazyflie quadcopters demonstrate the ability of Def-MARL to safely coordinate agents to complete complex collaborative tasks compared to other methods.

  PublisherOA PDFDOI

• Structured Interfaces for Automated Reasoning with 3D Scene Graphs

  ArXiv.org · 2025-10-18

  preprintOpen access

  In order to provide a robot with the ability to understand and react to a user's natural language inputs, the natural language must be connected to the robot's underlying representations of the world. Recently, large language models (LLMs) and 3D scene graphs (3DSGs) have become a popular choice for grounding natural language and representing the world. In this work, we address the challenge of using LLMs with 3DSGs to ground natural language. Existing methods encode the scene graph as serialized text within the LLM's context window, but this encoding does not scale to large or rich 3DSGs. Instead, we propose to use a form of Retrieval Augmented Generation to select a subset of the 3DSG relevant to the task. We encode a 3DSG in a graph database and provide a query language interface (Cypher) as a tool to the LLM with which it can retrieve relevant data for language grounding. We evaluate our approach on instruction following and scene question-answering tasks and compare against baseline context window and code generation methods. Our results show that using Cypher as an interface to 3D scene graphs scales significantly better to large, rich graphs on both local and cloud-based models. This leads to large performance improvements in grounded language tasks while also substantially reducing the token count of the scene graph content. A video supplement is available at https://www.youtube.com/watch?v=zY_YI9giZSA.

  PublisherOA PDFDOI

• Code-as-Symbolic-Planner: Foundation Model-Based Robot Planning via Symbolic Code Generation

  2025-10-19 · 2 citations

  articleSenior author

  Recent works have shown great potential of Large Language Models (LLMs) in robot task and motion planning (TAMP). Current LLM approaches generate text- or code-based reasoning chains with sub-goals and action plans. However, they do not fully leverage LLMs’ symbolic computing and code generation capabilities. Many robot TAMP tasks involve complex optimization under multiple constraints, where pure textual reasoning is insufficient. While augmenting LLMs with predefined solvers and planners improves performance, it lacks generalization across tasks. Given LLMs’ growing coding proficiency, we enhance their TAMP capabilities by steering them to generate code as symbolic planners for optimization and constraint verification. Unlike prior work that uses code to interface with robot action modules or pre-designed planners, we steer LLMs to generate code as solvers, planners, and checkers for TAMP tasks requiring symbolic computing, while still leveraging textual reasoning to incorporate common sense. With a multi-round guidance and answer evolution framework, the proposed Code-as-Symbolic-Planner improves success rates by average 24.1% over best baseline methods across seven typical TAMP tasks and three popular LLMs. Code-as-Symbolic-Planner shows strong effectiveness and generalizability across discrete and continuous environments, 2D/3D simulations and real-world settings, as well as single- and multi-robot tasks with diverse requirements. See our project website<sup xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">†</sup> for prompts, videos, and code.

  PublisherDOI

Recent grants

• CAREER: DeepCertify: Data-driven Formal Approach to Safe Autonomy

  NSF · $583k · 2023–2028

Frequent coauthors

• Charles Dawson

  Massachusetts Institute of Technology

  50 shared

• Sayan Mitra

  University of Illinois Urbana-Champaign

  36 shared

• Oswin So

  15 shared

• Zengyi Qin

  Tsinghua University

  14 shared

• Yue Meng

  American Institute of Aeronautics and Astronautics

  13 shared

• Songyuan Zhang

  13 shared

• Songyuan Zhang

  10 shared

• Kunal Garg

  Massachusetts Institute of Technology

  10 shared

Awards & honors

• NSF CAREER Award (2023)
• AFOSR Young Investigator Program (YIP) Award (2023)
• ONR Young Investigator Program (YIP) Award (2025)
• Innovators under 35 by MIT Technology Review (2021)
• ACM Doctoral Dissertation Award (2020)