Stephanie Gil
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Previous Organizations / Institutes
The organizations / institutes from the scholar's other articles in the past ten years
Research Subjects
Stephanie Gil has a total of 53 co-authors. Here are the top co-authors by number of shared publications:
- Andrea Goldsmith
Stony Brook University, Stony Brook, NY, USA
11 shared publications - Angelia Nedić
School of Electrical, Computer and Energy Engineering, Arizona State University, Tempe, USA
11 shared publications - Michal Yemini
Faculty of Engineering, Bar-Ilan University, Ramat Gan, Israel
9 shared publications - Matthew Cavorsi
School of Engineering and Applied Sciences, Harvard University, Cambridge, MA, USA
6 shared publications - Orhan Eren Akgün
School of Engineering and Applied Sciences Harvard University 02138. Cambridge MA
5 shared publications - Dimitri P. Bertsekas
M.I.T.
3 shared publications
Latest Publications
- 1 September 2026
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Robotics and Automation Letters
- Vol. 11 (9) , 10983-10990
WiSER-X: Wireless Signal-Based Decentralized Multi-Robot Exploration Without Explicit Information Exchange
- 1 September 2026
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Robotics and Automation Letters
- Vol. 11 (9) , 10983-10990
WiSER-X: Wireless Signal-Based Decentralized Multi-Robot Exploration Without Explicit Information Exchange
- 26 March 2026
- Published by Institute of Electrical and Electronics Engineers (IEEE) in Proceedings of the IEEE
- Vol. 113 (10) , 1121-1154
How Physicality Enables Cy-Trust: A New Era of Trust-Centered Cyber–Physical Systems
Cyber–physical multiagent systems are driving rapid technological advancements that automate a wide range of critical functions, thereby enabling safer, more accessible, and more efficient autonomous operations across diverse sectors. We refer to the capability of such systems to self-organize and coordinate toward accomplishing shared objectives as autonomy. The unique characteristics of these systems prompt a reevaluation of their security concepts, including their vulnerabilities, and mechanisms to mitigate these vulnerabilities. This survey article examines how advancements in wireless networking, coupled with sensing and computing capabilities, can foster novel security concepts for autonomous cyber–physical systems (CPSs). It delves into three main themes related to securing multiagent CPSs. First, we discuss the threats that are particularly relevant to multiagent CPSs, given the potential lack of trustworthiness between agents. Second, we present prospects for sensing, contextual awareness, and authentication, enabling the inference and measurement of a form of interagent “quantitative trust” or “cy-trust” for these systems. Third, we elaborate on the application of quantifiable trust notions to enable “resilient coordination,” where “resilient” signifies sustained functionality amid attacks on multiagent CPSs. This survey unveils the cyber–physical character of future interconnected systems as a pivotal catalyst for realizing robust autonomy.
- 26 March 2026
- Published by Institute of Electrical and Electronics Engineers (IEEE) in Proceedings of the IEEE
- Vol. 113 (10) , 1121-1154
How Physicality Enables Cy-Trust: A New Era of Trust-Centered Cyber–Physical Systems
Cyber–physical multiagent systems are driving rapid technological advancements that automate a wide range of critical functions, thereby enabling safer, more accessible, and more efficient autonomous operations across diverse sectors. We refer to the capability of such systems to self-organize and coordinate toward accomplishing shared objectives as autonomy. The unique characteristics of these systems prompt a reevaluation of their security concepts, including their vulnerabilities, and mechanisms to mitigate these vulnerabilities. This survey article examines how advancements in wireless networking, coupled with sensing and computing capabilities, can foster novel security concepts for autonomous cyber–physical systems (CPSs). It delves into three main themes related to securing multiagent CPSs. First, we discuss the threats that are particularly relevant to multiagent CPSs, given the potential lack of trustworthiness between agents. Second, we present prospects for sensing, contextual awareness, and authentication, enabling the inference and measurement of a form of interagent “quantitative trust” or “cy-trust” for these systems. Third, we elaborate on the application of quantifiable trust notions to enable “resilient coordination,” where “resilient” signifies sustained functionality amid attacks on multiagent CPSs. This survey unveils the cyber–physical character of future interconnected systems as a pivotal catalyst for realizing robust autonomy.
- 18 December 2025
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Robotics
- Vol. 42 (15523098) , 1128-1129
Corrections to “Characterizing Trust and Resilience in Distributed Consensus for Cyberphysical Systems”
In this correspondence, we correct the following points in the above paper.
- 18 December 2025
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Robotics
- Vol. 42 (15523098) , 1128-1129
Corrections to “Characterizing Trust and Resilience in Distributed Consensus for Cyberphysical Systems”
In this correspondence, we correct the following points in the above paper.
- 12 September 2025
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Robotics
- Vol. 41 (15523098) , 5533-5549
quasi-Dynamic Crowd Vetting: Collaborative Detection of Malicious Robots in Dynamic Communication Networks
We are interested in the problem where robots traverse through an environment modeled by a graph of discrete sites, and an unknown subset of the multirobot team is malicious. Previous works require that each robot gathers information about the trustworthiness of all other robots, called trust observations, which can be time consuming in large networks. This article decreases the time required to estimate trustworthiness by building upon an algorithm that leverages the concept of “crowd vetting” and the opinion of trusted neighbors. This allows each robot to estimate trust in dynamic scenarios, where the team size, robot neighborhoods, and robot legitimacy can change. In particular, we employ an assumption that there exists quasi-dynamic time periods, where if a robot’s legitimacy remains fixed for a sufficient length of time, its trustworthiness can be characterized. In this setting, we develop a closed-form expression for the critical number of time-steps required for our algorithm to successfully identify the true legitimacy of each robot within a specified failure probability. We show that the number of time-steps required for robots to correctly estimate the trust of all other robots increases logarithmically with the number of robots when robots do not leverage neighboring opinions, called the direct protocol. Conversely, for most general graph topologies, the number of time-steps required remains constant as the number of robots increases when our proposed algorithm, called quasi-dynamic crowd vetting (DCV), is used, for a fixed ratio of legitimate to malicious robots. Finally, our theoretical results are successfully validated through simulated persistent surveillance tasks where robots maintain a desired distribution of robots over sites in the environment.
- 12 September 2025
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Robotics
- Vol. 41 (15523098) , 5533-5549
quasi-Dynamic Crowd Vetting: Collaborative Detection of Malicious Robots in Dynamic Communication Networks
We are interested in the problem where robots traverse through an environment modeled by a graph of discrete sites, and an unknown subset of the multirobot team is malicious. Previous works require that each robot gathers information about the trustworthiness of all other robots, called trust observations, which can be time consuming in large networks. This article decreases the time required to estimate trustworthiness by building upon an algorithm that leverages the concept of “crowd vetting” and the opinion of trusted neighbors. This allows each robot to estimate trust in dynamic scenarios, where the team size, robot neighborhoods, and robot legitimacy can change. In particular, we employ an assumption that there exists quasi-dynamic time periods, where if a robot’s legitimacy remains fixed for a sufficient length of time, its trustworthiness can be characterized. In this setting, we develop a closed-form expression for the critical number of time-steps required for our algorithm to successfully identify the true legitimacy of each robot within a specified failure probability. We show that the number of time-steps required for robots to correctly estimate the trust of all other robots increases logarithmically with the number of robots when robots do not leverage neighboring opinions, called the direct protocol. Conversely, for most general graph topologies, the number of time-steps required remains constant as the number of robots increases when our proposed algorithm, called quasi-dynamic crowd vetting (DCV), is used, for a fixed ratio of legitimate to malicious robots. Finally, our theoretical results are successfully validated through simulated persistent surveillance tasks where robots maintain a desired distribution of robots over sites in the environment.
- 20 June 2025
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Control Systems Letters
- Vol. 9, 1544-1549
Friedkin-Johnsen Model is Distributed Gradient Descent
The Friedkin-Johnsen (FJ) model describes how agents adjust their opinions through repeated interactions while accounting for the influence of agents who are partially stubborn. In this work, we demonstrate that the FJ model is stepwise equivalent to solving the average consensus problem via distributed gradient descent. This perspective provides a unifying framework that bridges opinion dynamics and optimization, enabling the application of well-established results from the optimization literature. To illustrate this, we examine the recently proposed FJ model with diminishing stubbornness and extend prior results that were concerned with fixed communication graphs to time-varying and jointly connected communication graphs. We derive convergence guarantees and analyze convergence rates under these relaxed assumptions. Finally, we present numerical experiments on random graphs to showcase the impact of diminishing stubbornness dynamics on convergence in both static and time-varying settings.
- 20 June 2025
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Control Systems Letters
- Vol. 9, 1544-1549
Friedkin-Johnsen Model is Distributed Gradient Descent
The Friedkin-Johnsen (FJ) model describes how agents adjust their opinions through repeated interactions while accounting for the influence of agents who are partially stubborn. In this work, we demonstrate that the FJ model is stepwise equivalent to solving the average consensus problem via distributed gradient descent. This perspective provides a unifying framework that bridges opinion dynamics and optimization, enabling the application of well-established results from the optimization literature. To illustrate this, we examine the recently proposed FJ model with diminishing stubbornness and extend prior results that were concerned with fixed communication graphs to time-varying and jointly connected communication graphs. We derive convergence guarantees and analyze convergence rates under these relaxed assumptions. Finally, we present numerical experiments on random graphs to showcase the impact of diminishing stubbornness dynamics on convergence in both static and time-varying settings.
- 10 June 2025
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Automatic Control
- Vol. 70 (12) , 7788-7801
Learning Trust Over Directed Graphs in the Presence of Adversaries
We investigate the problem of learning the trustworthiness of agents in a directed graph, where agents receive stochastic trust information about their in-neighbors. We consider a setting where an unknown subset of agents may be malicious and disseminate misinformation about the trustworthiness of other agents. We present a learning protocol where agents determine their trusted neighborhoods based on the stochastic trust information they receive, and leverage the opinions of their trusted neighbors to assess the trustworthiness of other agents. We show that, under mild assumptions about the communication network, agents can learn the trustworthiness of other agents almost surely. Additionally, we establish an expected geometric convergence rate of the algorithm based on the properties of the stochastic trust values and the communication graph. Finally, we present numerical studies demonstrating that our convergence results hold in practice across various network topologies and with different numbers of malicious agents in the network.
- 10 June 2025
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Automatic Control
- Vol. 70 (12) , 7788-7801
Learning Trust Over Directed Graphs in the Presence of Adversaries
We investigate the problem of learning the trustworthiness of agents in a directed graph, where agents receive stochastic trust information about their in-neighbors. We consider a setting where an unknown subset of agents may be malicious and disseminate misinformation about the trustworthiness of other agents. We present a learning protocol where agents determine their trusted neighborhoods based on the stochastic trust information they receive, and leverage the opinions of their trusted neighbors to assess the trustworthiness of other agents. We show that, under mild assumptions about the communication network, agents can learn the trustworthiness of other agents almost surely. Additionally, we establish an expected geometric convergence rate of the algorithm based on the properties of the stochastic trust values and the communication graph. Finally, we present numerical studies demonstrating that our convergence results hold in practice across various network topologies and with different numbers of malicious agents in the network.
- Open Access
Multi-Agent Trustworthy Consensus under Random Dynamic Attacks
In this work, we study the consensus problem in which legitimate agents send their values over an undirected communication network in the presence of an unknown subset of malicious or faulty agents. In contrast to former works, we generalize and characterize the properties of consensus dynamics with dependent sequences of malicious transmissions with dynamic (time-varying) rates, based on not necessarily independent trust observations. We consider a detection algorithm utilizing stochastic trust observations available to legitimate agents. Under these conditions, legitimate agents almost surely classify their neighbors and form their trusted neighborhoods correctly with decaying misclassification probabilities. We further prove that the consensus process converges almost surely despite the existence of malicious agents. For a given value of failure probability, we characterize the deviation from the nominal consensus value ideally occurring when there are no malicious agents in the system. We also examine the convergence rate of the process in finite time. Numerical simulations show the convergence among agents and indicate the deviation under different attack scenarios.
Open AccessMulti-Agent Trustworthy Consensus under Random Dynamic Attacks
In this work, we study the consensus problem in which legitimate agents send their values over an undirected communication network in the presence of an unknown subset of malicious or faulty agents. In contrast to former works, we generalize and characterize the properties of consensus dynamics with dependent sequences of malicious transmissions with dynamic (time-varying) rates, based on not necessarily independent trust observations. We consider a detection algorithm utilizing stochastic trust observations available to legitimate agents. Under these conditions, legitimate agents almost surely classify their neighbors and form their trusted neighborhoods correctly with decaying misclassification probabilities. We further prove that the consensus process converges almost surely despite the existence of malicious agents. For a given value of failure probability, we characterize the deviation from the nominal consensus value ideally occurring when there are no malicious agents in the system. We also examine the convergence rate of the process in finite time. Numerical simulations show the convergence among agents and indicate the deviation under different attack scenarios.
- Open Access
Pro-Routing: Proactive Routing of Autonomous Multi-Capacity Robots for Pickup-and-Delivery Tasks
We consider a multi-robot setting, where we have a fleet of multi-capacity autonomous robots that must service spatially distributed pickup-and-delivery requests with fixed maximum wait times. Requests can be either scheduled ahead of time or they can enter the system in real-time. In this setting, stability for a routing policy is defined as the cost of the policy being uniformly bounded over time. Most previous work either solve the problem offline to theoretically maintain stability or they consider dynamically arriving requests at the expense of the theoretical guarantees on stability. In this paper, we aim to bridge this gap by proposing a novel proactive rollout-based routing framework that adapts to real-time demand while still provably maintaining the stability of the learned routing policy. We derive provable stability guarantees for our method by proposing a fleet sizing algorithm that obtains a sufficiently large fleet that ensures stability by construction. To validate our theoretical results, we consider a case study on real ride requests for Harvard's evening Van System. We also evaluate the performance of our framework using the currently deployed smaller fleet size. In this smaller setup, we compare against the currently deployed routing algorithm, greedy heuristics, and Monte-Carlo-Tree-Search-based algorithms. Our empirical results show that our framework maintains stability when we use the sufficiently large fleet size found in our theoretical results. For the smaller currently deployed fleet size, our method services 6% more requests than the closest baseline while reducing median passenger wait times by 33%.
Open AccessPro-Routing: Proactive Routing of Autonomous Multi-Capacity Robots for Pickup-and-Delivery Tasks
We consider a multi-robot setting, where we have a fleet of multi-capacity autonomous robots that must service spatially distributed pickup-and-delivery requests with fixed maximum wait times. Requests can be either scheduled ahead of time or they can enter the system in real-time. In this setting, stability for a routing policy is defined as the cost of the policy being uniformly bounded over time. Most previous work either solve the problem offline to theoretically maintain stability or they consider dynamically arriving requests at the expense of the theoretical guarantees on stability. In this paper, we aim to bridge this gap by proposing a novel proactive rollout-based routing framework that adapts to real-time demand while still provably maintaining the stability of the learned routing policy. We derive provable stability guarantees for our method by proposing a fleet sizing algorithm that obtains a sufficiently large fleet that ensures stability by construction. To validate our theoretical results, we consider a case study on real ride requests for Harvard's evening Van System. We also evaluate the performance of our framework using the currently deployed smaller fleet size. In this smaller setup, we compare against the currently deployed routing algorithm, greedy heuristics, and Monte-Carlo-Tree-Search-based algorithms. Our empirical results show that our framework maintains stability when we use the sufficiently large fleet size found in our theoretical results. For the smaller currently deployed fleet size, our method services 6% more requests than the closest baseline while reducing median passenger wait times by 33%.
- Top Cited Papers
- 22 January 2025
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Automatic Control
- Vol. 70 (6) , 3952-3967
Resilient Distributed Optimization for Multiagent Cyberphysical Systems
This work focuses on the problem of distributed optimization in multi-agent cyberphysical systems, where a legitimate agent's iterates are influenced both by the values it receives from potentially malicious neighboring agents, and by its own self-serving target function. We develop a new algorithmic and analytical framework to achieve resilience for the class of problems where stochastic values of trust between agents exist and can be exploited. In this case, we show that convergence to the true global optimal point can be recovered, both in mean and almost surely, even in the presence of malicious agents. Furthermore, we provide expected convergence rate guarantees in the form of upper bounds on the expected squared distance to the optimal value. Finally, numerical results are presented that validate our analytical convergence guarantees even when the malicious agents compose the majority of agents in the network and where existing methods fail to converge to the optimal nominal points.
- 22 January 2025
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Automatic Control
- Vol. 70 (6) , 3952-3967
Top Cited PapersResilient Distributed Optimization for Multiagent Cyberphysical Systems
This work focuses on the problem of distributed optimization in multi-agent cyberphysical systems, where a legitimate agent's iterates are influenced both by the values it receives from potentially malicious neighboring agents, and by its own self-serving target function. We develop a new algorithmic and analytical framework to achieve resilience for the class of problems where stochastic values of trust between agents exist and can be exploited. In this case, we show that convergence to the true global optimal point can be recovered, both in mean and almost surely, even in the presence of malicious agents. Furthermore, we provide expected convergence rate guarantees in the form of upper bounds on the expected squared distance to the optimal value. Finally, numerical results are presented that validate our analytical convergence guarantees even when the malicious agents compose the majority of agents in the network and where existing methods fail to converge to the optimal nominal points.
- 2 December 2024
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Control Systems Letters
- Vol. 8, 2679-2684
Friedkin-Johnsen Model With Diminishing Competition
This letter studies the Friedkin-Johnsen (FJ) model with diminishing competition, or stubbornness. The original FJ model assumes that each agent assigns a constant competition weight to its initial opinion. In contrast, we investigate the effect of diminishing competition on the convergence point and speed of the FJ dynamics. We prove that, if the competition is uniform across agents and vanishes asymptotically, the convergence point coincides with the nominal consensus reached with no competition. However, the diminishing competition slows down convergence according to its own rate of decay. We study this phenomenon analytically and provide upper and lower bounds on the convergence rate. Further, if competition is not uniform across agents, we show that the convergence point may not coincide with the nominal consensus point. Finally, we evaluate our analytical insights numerically.
- 2 December 2024
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Control Systems Letters
- Vol. 8, 2679-2684
Friedkin-Johnsen Model With Diminishing Competition
This letter studies the Friedkin-Johnsen (FJ) model with diminishing competition, or stubbornness. The original FJ model assumes that each agent assigns a constant competition weight to its initial opinion. In contrast, we investigate the effect of diminishing competition on the convergence point and speed of the FJ dynamics. We prove that, if the competition is uniform across agents and vanishes asymptotically, the convergence point coincides with the nominal consensus reached with no competition. However, the diminishing competition slows down convergence according to its own rate of decay. We study this phenomenon analytically and provide upper and lower bounds on the convergence rate. Further, if competition is not uniform across agents, we show that the convergence point may not coincide with the nominal consensus point. Finally, we evaluate our analytical insights numerically.