description of image


Welcome! I am an associate professor at the University of Maryland, College Park in Aerospace Engineering with a joint appointment at the Institute for Systems Research. My work focuses on formal methods for multi-agent autonomous systems. Browse around for more information.


I received my Ph.D. in Mechanical Engineering from the California Institute of Technology in June 2013. Before that, I graduated from Harvard University with my S.B. in Mechanical Engineering and Material Science in 2007. My dissertation focused on using formal methods and specification languages in the design and analysis of large-scale, complex, distributed sensing, actuation, and control systems with an emphasis on correct-by-construction control synthesis.

My research interests comprise the areas of control and dynamical systems, optimization, and formal methods with applications in autonomy, planning, and system identification. In particular, I am interested in developing and designing robust cyber-physical systems that are capable of performing high-level, complex goals while reacting to dynamic and possibly adversarial environments. The overall research goal is to provide provably correct results through the use of three key aspects from my research: (1) rigorous theory drawn from mathematics, controls, and computer science (e.g., graph theory and formal methods); (2) computational tractability and how complexity may be overcome through decompositions and interface contracts; and (3) applications to real-world/industry problems (e.g., smart-grid, autonomous vehicles, security of cyber-physical systems).



Cooperative Re-Planning via Sensor Inference and Geospatial Data Integration

The long-term goal of this research is to enable a heterogenous team of UAS to autonomously deploy sensors into a partially known environment. This project will use information contained from an ARL-developed Geospatial Data Integration Server (GDIS) to infer best locations for initial deployment (i.e., locations utilizing a minimal number of sensors for maximum information gain). As deployed sensors begin relaying data, updates can be made for relocation of sensors to better enhance knowledge for terrain classification, terrain traversability analysis for ground vehicles, target localization, or cooperative planning of the UAS.

Integrated Power-Platform Innovations for Hybrid-Electric Rotorcraft

This VLRCOE task is a joint project with Professor Anubhav Datta. The main objective is to advance hybrid-electric rotorcraft propulsion through power and platform innovations in controlled power-sharing, test-bed development and data-acquisition, and modeling and simulation.

Development of an Impact Resistant Swept Wing

In this project, the experience gained by ARL researchers, coupled with the prior work of UMD researchers, will be exploited to develop a flyable swept wing mechanism that achieves two key goals. The first goal will be enable the main lifting wing of a fixed-wing uncrewed air system (UAS) to accept unanticipated leading edge strikes from obstacles, to deflect aft to reduce the shock loads associated with such leading edge strikes, and to return the wing to its steady level flight configuration via an actuation system (either passive or active). The second goal will seek to utilize the actuation system in the swept wing mechanism, in conjunction with the collision avoidance system to be developed in subsequent tasks, to reduce the frontal area of the wing to avoid impacts by sweeping the wings aft to avoid such impacts.

AUVSI SUAS Competition

The Student Unmanned Aircraft Systems (SUAS) Competition, aimed at stimulating and fostering interest in unmanned air systems, technologies and careers, focuses on engaging students in a challenging mission. It requires the design, integration, and demonstration of a system capable of conducting air operations, which include autonomous flight, navigation of a specified course, and use of onboard payload sensors.

Previous Projects

Autonomous Drone Racing

Competitive autonomous drone racing is an engineering and computer science challenge that requires an understanding of computer vision, the ability to develop algorithms that incorporate the gate detection, and programming logic for the drone to understand when it has completed tasks. Students compete annually at the IEEE International Conference on Intelligent Robots and Systems (IROS).

Real-Time Flight Planning Simulation Software for Unmanned Aircraft Systems/Millennium Engineering and Integration

This project aims to develop and demonstrate a real-time flight simulation tool that integrates mission planning, ground command/control, onboard situational awareness, and onboard re-planning for unmanned aircraft vehicles (UAV). The tool will aid in the assessment of potential flight-safety hardware subsystems and embedded algorithmic solutions to enable operation of UAVs within the National Airspace (NAS) and uncontrolled urban airspace.

Electric VTOL Real-Time HIL Test Environment/HopFlyt

This project develops preliminary flight control logic in a simulated environment for use in flight testing a proof of concept UAV that aims to “hop” over traffic for distances up to 30 miles, carrying up to three passengers at a time. The project envisions reducing the 90-minute drive from Baltimore to Washington, D.C. during rush hour to just 15 minutes.

Planning and Metareasoning for Multi-agent Systems with Variable Communication Availability/AFRL

Joint project with Jeffrey Hermann (ME), Shapour Azarm (ME), and Michael Otte (AE). We generate knowledge about multi-agent planning and metareasoning based on dynamic tasking and communication availability. The proposed research includes two key areas: (1) develop, understand, and characterize the performance of hybrid centralized/decentralized multi-agent planning methods with dynamic tasking for different levels of communication availability; and (2) develop, analyze, and compare metareasoning techniques that enable autonomous agents in multi-agent systems to select the most appropriate planning method based on the real-time needs for tasking (including task removal during execution) and observations of the current level of communication availability.

Route Planning With Complex Geometries/NAVAIR

We explore a methodology for training recurrent neural networks in replicating path planning solutions from optimization problems. Training data is generated from a kino-dynamic rapidly-exploring random tree, from which a recurrent neural network is trained upon to produce the state path through fixed time-step execution. Path-tracking controllers are formulated to follow the path generated by the network alongside the use of local potential functions to mitigate minor constraint violations. The control signal from such a controller should mimic that of the optimized solution, but can be generated orders of magnitude faster than the slower planning solutions.

Robust Semi-Autonomous Swarm Tactics for Situational Awareness in Uncertain Environments/DARPA

Joint Project with Ming Lin (CS), Dinesh Manocha (CS), and Michael Otte (AE). This work creates novel and useful semi-autonomous aerial swarm tactics for the surveillanceof a hostile urban environment with uncertainty by enabling a 3D interactive model of the urbanenvironment to be created and continually updated with visual and geometric data captured bythe swarm. These tactics will be robust with respect to uncertainty in the environment as wellas sensor noise. This will enable a single human user to have real-time battlefield situationalawareness within dynamic hostile environments. While current use of unmanned systems requiresmultiple humans to operate a single vehicle, this capability would allow a single operator to controlmultiple vehicles.

Mission planning for autonomous systems in dynamic environments using open-source autopilots/NAVAIR

This project investigates the integration and certification of open source autopilots in unmanned aerial vehicles using motion planning as a formal framework. The objective is for UMD to develop autopilot architectures and mission planning algorithms using open-source software and off-the-shelf autopilots (e.g., Navio2).

Non-Cooperative Detect and Avoid Capabilities for UAS Platforms/Lockheed Martin

This project leverages University of Maryland research in the areas of UAS detection for counter-UAS applications and autonomous vehicle path and trajectory planning to address non-cooperative detect and avoid capability for UAS systems.

Cybersecurity Detection for UAVs/MITRE

The main goal of this project is to provide theoretical guarantees on the ability to estimate and recover from security threats to cyberphysical systems. We aim to develop a simulation and hardware testbed in which the proposed mathematical framework can be used to detect, assess, and recover from threats (both internal and external).


*Updated May 2023

Journal Publications

  1. [J.18] Generating Certification Evidence for the Certification of Collision Avoidance in Autonomous Surface Vessels. D. Costello and H. Xu. In: Journal of Maritime Policy and Management. 2023. [Link]
  2. [J.17] Using a Run-Time Assurance Approach for Certifying Autonomy Within Naval Aviation. D. Costello and H. Xu. In: Systems Engineering. 2023. [Link]
  3. [J.16] Distributed Task Allocation Algorithms for Multi-Agent Systems with Very Low Communication. A. Bapat et al. In: IEEE Access. 2022. [Link]
  4. [J.15] A Path-Dependant Approach for Characterizing the Legal Governance of Autonomous Systems. J.E. Borson and H. Xu. In: IEEE Access. 2022. [Link]
  5. [J.14] Communication-Aware Multi-Agent Metareasoning for Decentralized Task Allocation. E. Carrillo et al. In: IEEE Access. 2021. [Link]
  6. [J.13] Relating Sensor Degradation to Vehicle Situational Awareness for Autonomous Air Vehicle Certification. D. Costello and H. Xu. In: AIAA Journal of Aerospace Information Systems. 2021. [Link]
  7. [J.12] Robust Nonlinear Control-Based Trajectory Tracking for Quadrotors Under Uncertainty. K. Kidambi, F. Fermuller, Y. Alomoinos, and H. Xu. In: IEEE Control Systems Letters. 2020. [Link]
  8. [J.11] Autonomous Flight Test Data in Support of a Safety of Flight Certification. D. Costello and H. Xu. In: AIAA Journal of Air Transportation. 2020. [Link]
  9. [J.10] Generating Certification Evidence for Autonomous Aerial Vehicles Decision Making. D. Costello and H. Xu. In: AIAA Journal of Aerospace Information Systems. 2020. [Link]
  10. [J.9] Metareasoning Structures, Problems, and Modes for Multiagent Systems: A Survey. S. Langlois et al. In: IEEE Access. 2020. [Link]
  11. [J.8] Mitigation of Ground Impact Hazard for Safe Unmanned Aerial Vehicle Operations. A. Poissant, L.Castano, H. Xu. In: AIAA Journal of Aerospace Information Systems. 2020. [Link]
  12. [J.7] Experimental Comparison of Decentralized Task Allocation Algorithms Under Imperfect Communication. Nayak, S., et al. In: IEEE Robotics and Automation Letters. 2020. [Link]
  13. [J.6] LSwarm: Efficient Collision Avoidance for Large Swarms with Coverage Constraints in Complex Urban Scenes. S. Arul, A. Sathyamoorthy, S. Patel, M. Otte, H. Xu, M. Lin, and D. Manocha. In: IEEE Robotics and Automation Letters. 2019. [Link]
  14. [J.5] Fast 3D Collision Avoidance Algorithm for Fixed-Wing UAS. Z. Lin, E. Mortimer, L. Castano, H. Xu. In: Journal of Intelligent and Robotic Systems. 2019. [Link]
  15. [J.4] Hierarchal Application of Receding Horizon Synthesis and Dynamic Allocation for UAVs Fighting Fires. J. Shaffer, E. Carillo, and H. Xu. In: IEEE Access. 2018. [Link]
  16. [J.3] Specification and Synthesis for Aircraft Electric Power Distribution. H. Xu, U. Topcu, and R. M. Murray. In: IEEE Transactions on Networked Control Systems. 2015. [Link]
  17. [J.2] Control Software Synthesis for A Vehicular Electric Power Distribution Testbed. R. Rogersten, H. Xu, N. Ozay, U. Topcu, and R. M. Murray. In: Journal of Aerospace Information Systems. 11:10, 665-678 2015. [Link]
  18. [J.1] A Contract-Based Methodology for Aircraft Electric Power System Design. P. Nuzzo, H. Xu, N. Ozay, R. M. Murray, A. Sangiovanni-Vincentelli, et al. In: IEEE Access. 2013. [Link]

Conference Publications

  1. [C.23] Fuzzy Logic and Mahalanobis Distance Algorithms for Fault Detection in Fixed-Wing UAVs. R. Gomez, L. Castano, and H. Xu. In: AIAA Aviation Forum. 2023. [Preprint]
  2. [C.22] Multi-Agent Ergodic Coverage in Urban Environments. S. Patel et al. In: IEEE International Conference on Robotics and Automation (ICRA). 2021. [Link]
  3. [C.21] UAV Collision Avoidance with Varying Trigger Time. Z. Lin, L. Castano, and H. Xu. In: IEEE International Conference on Unmanned Aircraft Systems. 2020. [Link]
  4. [C.20] Decentralized Task Allocation in Multi-Agent Systems Using a Decentralized Genetic Algorithm. R. Patel et al. In: IEEE International Conference on Robotics and Automation (ICRA). 2020. [Link]
  5. [C.19] Monitoring Access to User Defined Areas with Swarms of UAVs in Urban Environments. M. Gupta, M. Lin, D. Manocha, H. Xu, M. Otte. In: Multi-Robot and Multi-Agent Systems. 2019. [Link]
  6. [C.18] Expanding Kino-dynamic Rapidly-exploring Random Trees with Recurrent Neural Networks. J. Shaffer and H. Xu.. In: IEEE Conference on Decision and Control. 2019. [Link]
  7. [C.17] Safe decision making for risk mitigation of UAS. L. Castano and H. Xu. In: International Conference on Unmanned Aircraft Systems. 2019. [Link]
  8. [C.16] Formal methods and neural networks for specification of autonomous swarm behavior. J. Shaffer and H. Xu. In: SPIE Defense and Commercial Sensing. April 2019. Invited Paper [Link]
  9. [C.15] Determination of Safe Landing Zones for an Autonomous UAS using Elevation and Population Density Data. E. Carney, L. Castano, and H. Xu In: AIAA SciTech, San Diego, CA. January 2019. [Link]
  10. [C.14] The Future of Legal and Ethical Regulations for Autonomous Robotics. H. Xu, J.Borson. In: IEEE/RSJ Intelligent Robots and Systems, Madrid, Spain. October 2018. [Link]
  11. [C.13] Receding Horizon Synthesis and Dynamic Allocation of UAVs to Fight Fires. J. Shaffer, E. Carrillo, H. Xu. In: IEEE Conference on Control Technology and Applications, Copenhagen, Denmark. August 2018. [Link]
  12. [C.12] A Fast Obstacle Collision Avoidance Algorithm for Fixed Wing UAS. Z. Lin, L. Castano, H. Xu. In: International Conference on Unmanned Aircraft Systems, Dallas, TX. June 2018. [Link]
  13. [C11] Ground Impact and Hazard Mitigation for Safer UAV Flight Response. A. Poissant, L. Castano, H. Xu. In: International Conference on Unmanned Aircraft Systems, Dallas, TX. June 2018. [Link]
  14. [C.10] Communicating STEM: How podcasting can help women in STEM become better communicators. L. Claiborn and H. Xu. In: IEEE Women in Engineering Summit East, Baltimore, MD. December 2017. [Link]
  15. [C.9] A Systematic Approach to Mission and Scenario Planning for UAVs. N Shadab and H. Xu. In: IEEE Systems Conference (SysCon), 2016. [Link]
  16. [C.8] Security of Unmanned Aerial Vehicles: Dynamic State Estimation Under Cyber-Physical Attacks. L. Petnga and H. Xu In: IEEE Unmanned Aircraft Systems (ICUAS), 2016. [Link]
  17. [C.7] An Autonomous, Visually-Guided, Counter-sUAS Aerial Vehicle with Net Countermeasure. T. Horiuchi, et al. In: AIAA Atmospheric Flight Mechanics Conference, 2016. [Link]
  18. [C.6] Dynamic State Estimation in Distributed Aircraft Electric Control Systems via Adaptive Submodularity. Q. Maillet, H. Xu, N. Ozay, and R. M. Murray. In: IEEE Conference on Decision and Control, 2013. [Link]
  19. [C.5] From Formal Specifications to Software Models and Hardware Implementation of Reactive Protocols: An Aircraft Electric Power Testbed. R. Rogersten, N. Ozay, U. Topcu, H. Xu, and R. M. Murray. In: International Conference on Hybrid Systems: Computation and Control, 2013. [Link]
  20. [C.4] A Case Study on Reactive Protocols for Aircraft Electric Power Distribution H. Xu, U. Topcu, and R. M. Murray. In: IEEE Conference on Decision and Control, 2012. [Link]
  21. [C.3] TuLiP: A Software Toolbox for Receding Horizon Temporal Logic Planning. T. Wongpiromsarn, U. Topcu, N. Ozay, H. Xu and R. M. Murray. In: International Conference on Hybrid Systems: Computation and Control, 2011. [Link]
  22. [C.2] Load-shedding Probabilities of Power Systems with Renewable Power Generation and Energy Storage. H. Xu, U. Topcu, S. Low, C. Clarke, and K. M. Chandy. In: Allerton Conference on Communication, Control, and Computing, 2010. [Link]
  23. [C.1] A Simple Optimal Power Flow Model with Energy Storage. H. Xu, U. Topcu, S. Low, and K. M. Chandy. In: IEEE Conference on Decision and Control, 2010. [Link]

Please feel free to email about preprints.



  • John Schmidt, PhD candidate, AE
  • Sarah Wielgosz, PhD candidate, AE
  • Ruth Gomez Quezada, MS candidate, AE
  • Graham Buccheri, Undergraduate, CS
  • Jeriel Cortes, Undergraduate, AE


  • Estefany Carrillo, PhD 2021, AE. Thesis: Controller Synthesis and Formal Behavior Inference in Autonomous Systems
  • Zijie Lin, PhD 2021, ECE. Thesis: 3D Fast Geometric Collision Avoidance Algorithm (FGA)
  • Donald Costello, PhD 2020, ME. Thesis: Certifying an Autonomous System to Complete Tasks
  • Scott Morrison, MS 2023, Aerospace Engineering (Visiting Scholar)
  • Anshuman Singh, MS 2021, Systems Engineering
  • Prasheel Renkuntla, MS 2021, Robotics
  • Christopher Roth, MS 2020, Aerospace Engineering (Visting Scholar)
  • Swapneel Naphade, MS 2020, Robotics
  • Derek Thompson, MS 2020, Aerospace Engineering
  • Ed Carney, MS 2020, Aerospace Engineering
  • Suyash Yeotikar, MS 2020, Robotics
  • Joshua Shaffer, MS 2019, Aerospace Engineering
  • Vincenz Frenzel, MS 2019, Aerospace Engineering (Visiting Scholar)
  • Edward Mortimer, BS/MS 2019, Aerospace Engineering (Visiting Scholar)
  • Andrew Poissant, BS/MS 2018, Systems Engineering
  • Matthew Solomon, MS 2016, Aerospace Engineering
  • Niloofar Shadab, MS 2016, Systems Engineering
  • Bao Zhang, MS 2014, Systems Engineering Engineering


SUMMER 2023 SHORT COURSE: How to Build, Fly, and Certify Autonomous Aerial Systems

The University of Maryland (UMD) MATRIX Lab will offer our first short course on aerial autonomy in collaboration with the UMD UAS Research and Operations Center (UROC) this summer. Participants will learn about the certification process, plus verification and validation. They will get the opportunity to build and fly drones. This is just the first of several cutting-edge technology short courses we plan to offer at the University System of Maryland at Southern Maryland (USMSM) SMART Building.

FALL: ENAE 380 Flight Software Systems

Avionics using advanced sensor and computing technologies are at the heart of every modern aerospace vehicle. Advanced software systems improve cockpit safety and enable unmanned and deep-space missions. Object-oriented programming and software engineering concepts required to design and build complex flight software systems will be discussed. Other discussions include software validation, verification, real-time performance analysis to assess flight software system reliability and robustness; human-machine interfaces designed for piloted systems; automatic onboard data acquisition and decision-making for unmanned air and space vehicles.

SPRING: ENAE 450 Robotics Programming

Introduces students to the Robot Operating System (ROS) as well as to many of the available tools commonly used in robotics. Lectures focus on theory and structure, whereas laboratory sections will focus on applications and implementations. Students learn how to create software and simulations, interface to sensors and actuators, and integrate control algorithms. The course works through exercises involving a number of autonomous robots (i.e., ground and air vehicles) that students will eventually use in their subsequent RAS minor courses. Topics include: ROS architecture, console commands, ROS packages, simulation environments, visualizations, autonomous navigation, manipulation, and robot vision.

SPRING: ENES489p Special Topics in Engineering (Hands-On Systems Engineering Projects)

This hands-on design projects course will expose senior-level undergraduate and graduate-level students from all areas of engineering to exciting career opportunities in the systems engineering field. Students will be introduced to the technical aspects of systems engineering practice through team-based project development and a systematic step-by-step procedure for product development that includes working with a real-world customer to define operations concepts, requirements gathering and organization, synthesis of models of system behavior and system structure, functional allocation to create system design alternatives, formal assessment of design alternatives through tradeoff analysis, and established approaches to testing and validation/verification.

SPRING: ENSE 698B Applied Formal Methods

This course introduces best practices for the application of formal methods, a set of mathematically rigorous techniques for the formal specification, validation, and verification of safety-critical systems, of which aircraft and spacecraft are the prime examples. The course explores tools, techniques, and applications of formal methods, focusing on aerospace and robotic domains. Students examine the latest research to gain an understanding of the current state of the art, including the capabilities and limitations of applying formal methods for systems analysis.


3180 Glenn L. Martin Building
College Park, MD 20742
Phone: +1 301 405 1133