HRHossein Rastgoftar

Hossein Rastgoftar

*This is an algorithmically generated author record based on publicly available information.

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


Hossein Rastgoftar has a total of 22 co-authors. Here are the top co-authors by number of shared publications:


Latest Publications

  • Open Access

    Safety-Critical Reinforcement Learning with Viability-Based Action Shielding for Hypersonic Longitudinal Flight

    by Hossein Rastgoftar

    This paper presents a safety-critical reinforcement learning framework for nonlinear dynamical systems with continuous state and input spaces operating under explicit physical constraints. Hard safety constraints are enforced independently of the reward through action shielding and reachability-based admissible action sets, ensuring that unsafe behaviors are never intentionally selected during learning or execution. To capture nominal operation and recovery behavior within a single control architecture, the state space is partitioned into safe and unsafe regions based on membership in a safety box, and a mode-dependent reward is used to promote accurate tracking inside the safe region and recovery toward it when operating outside. To enable online tabular learning on continuous dynamics, a finite-state abstraction is constructed via state aggregation, and action selection and value updates are consistently restricted to admissible actions. The framework is demonstrated on a longitudinal point-mass hypersonic vehicle model with aerodynamic and propulsion couplings, using angle of attack and throttle as control inputs.

  • Open Access

    Finite-State Decentralized Policy-Based Control With Guaranteed Ground Coverage

    by Hossein Rastgoftar

    We propose a finite-state, decentralized decision and control framework for multi-agent ground coverage. The approach decomposes the problem into two coupled components: (i) the structural design of a deep neural network (DNN) induced by the reference configuration of the agents, and (ii) policy-based decentralized coverage control. Agents are classified as anchors and followers, yielding a generic and scalable communication architecture in which each follower interacts with exactly three in-neighbors from the preceding layer, forming an enclosing triangular communication structure. The DNN training weights implicitly encode the spatial configuration of the agent team, thereby providing a geometric representation of the environmental target set. Within this architecture, we formulate a computationally efficient decentralized Markov decision process (MDP) whose components are time-invariant except for a time-varying cost function defined by the deviation from the centroid of the target set contained within each agent communication triangle. By introducing the concept of Anyway Output Controllability (AOC), we assume each agent is AOC and establish decentralized convergence to a desired configuration that optimally represents the environmental target.

  • Open Access

    Deep Neural Network-Based Aerial Transport in the Presence of Cooperative and Uncooperative UAS

    by Muhammad Junayed Hasan Zahed, Hossein Rastgoftar

    We present a resilient deep neural network (DNN) framework for decentralized transport and coverage using uncrewed aerial systems (UAS) operating in $\mathbb{R}^n$. The proposed DNN-based mass-transport architecture constructs a layered inter-UAS communication graph from an initial formation, assigns time-varying communication weights through a forward scheduling mechanism that guides the team from the initial to the final configuration, and ensures stability and convergence of the resulting multi-agent transport dynamics. The framework is explicitly designed to remain robust in the presence of uncooperative agents that deviate from or refuse to follow the prescribed protocol. Our method preserves a fixed feed-forward topology but dynamically prunes edges to uncooperative agents, maintains convex, feedforward mentoring among cooperative agents, and computes global desired set points through a sparse linear relation consistent with leader references. The target set is abstracted by $N$ points that become final desired positions, enabling coverage-optimal transport while keeping computation low and guarantees intact. Extensive simulations demonstrate that, under full cooperation, all agents converge rapidly to the target zone with a 10\% boundary margin and under partial cooperation with uncooperative agents, the system maintains high convergence among cooperative agents with performance degradation localized near the disruptions, evidencing graceful resilience and scalability. These results confirm that forward-weight scheduling, hierarchical mentor--mentee coordination, and on-the-fly DNN restructuring yield robust, provably stable UAS transport in realistic fault scenarios.

  • VR-Based Control of Multicopter Operation

    by Jack T. Hughes, Garegin Mazmanyan, Mohammad Ghufran, Hossein Rastgoftar

    We present a VR-based teleoperation system for multirotor flight that renders a third-person view (TPV) of the vehicle together with a live 3D reconstruction of its surroundings. The system runs on an embedded GPU (Jetson Orin NX) with ROS2/WebXR integration and streams geometry and video to a headset for closed-loop control in pre viously unmapped spaces. We implement a first-person video (FPV) baseline and perform matched trials with two pilots in unmapped indoor spaces. Quantitative metrics are reported from repeated trials with one pilot (N = 8) TPV achieved task time comparable to FPV while improving proximal obstacle awareness (minimum obstacle distance +0.20m) and reducing contacts. These results indicate that TPV can preserve control quality while exposing hazards less visible in FPV, supporting safer teleoperation in unknown environments.

  • Open Access

    Experimental Validation of Decentralized Affine Transformation

    by Garegin Mazmanyan, Hossein Rastgoftar

    This paper presents an experimental validation of decentralized affine transformation (AT) in multi-agent systems using teams of mini-quadcopters. The AT framework enables an agent team to safely navigate constrained, obstacle-rich environments while allowing aggressive changes in inter-agent distances, which are formally characterized through the decomposition of the AT transformation matrix. Without loss of generality, we focus on two-dimensional AT, formulated as a decentralized leader-follower problem. In this formulation, three leader quadcopters are positioned at the vertices of a triangle, while all follower quadcopters remain within the triangle. The leaders know the desired trajectories prescribed by the AT, whereas the followers do not. Instead, the followers infer their trajectories through local communication governed by fixed communication weights determined by the initial spatial configuration of the team. Experimental results validate the asymptotic convergence of decentralized AT and demonstrate its capability to safely guide multi-agent teams through obstacle-laden environments.

  • Open Access

    Steering Opinion through Dynamic Stackelberg Optimization

    by Hossein Rastgoftar

    This paper employs the Friedkin-Johnsen (FJ) model to describe the dynamics of opinion evolution within a social network. Under the FJ framework, the society is divided into two subgroups that include stubborn agents and regular agents. The opinions of stubborn agents are not influenced by regular agents, whereas the opinions of regular agents evolve based on the opinions of their neighboring agents. By defining the origin as the desired collective opinion of the society, the objective of the paper is to minimize deviations from this desired opinion. To achieve this, a Stackelberg game is established between the stubborn and regular subgroups, where the opinion adjustments of the stubborn agents and the openness variables of regular agents serve as the decision variables. The proposed solution approach integrates quadratic programming and dynamic programming to optimize these decision variables at each discrete time step using forward and backward propagation.

  • Open Access

    Containment Control Approach for Steering Opinion in a Social Network

    by Hossein Rastgoftar

    The paper studies the problem of steering multi-dimensional opinion in a social network. Assuming the society of desire consists of stubborn and regular agents, stubborn agents are considered as leaders who specify the desired opinion distribution as a distributed reward or utility function. In this context, each regular agent is seen as a follower, updating its bias on the initial opinion and influence weights by averaging their observations of the rewards their influencers have received. Assuming random graphs with reducible and irreducible topology specify the influences on regular agents, opinion evolution is represented as a containment control problem in which stability and convergence to the final opinion are proven.

  • Safe Human-UAS Collaboration From High-Level Planning to Low-Level Tracking

    by Hossein Rastgoftar

    This paper studies the problem of safe human-uncrewed aerial system (UAS) collaboration in a shared work environment. By considering human and UAS as co-workers, we use Petri Nets to abstractly model evolution of shared tasks assigned to human and UAS co-workers. Particularly, the Petri Nets’ “places” represent work stations; and therefore, the Petri Nets’ transitions can formally specify displacements between the work stations. The paper’s first objective is to incorporate uncertainty regarding the intentions of human co-workers into motion planning for UAS, when UAS closely interacts with human co-workers. To this end, the proposed Petri Nets model uses “conflict” constructs to represent situations at which UAS deals with incomplete knowledge about human co-worker intention. The paper’s second objective is then to plan the motion of the UAS in a resilient and safe manner, in the presence of non-cooperative human co-workers. In order to achieve this objective, UAS equipped with onboard perception and decision-making capabilities are able to, through real-time processing of in-situ observation, predict human intention, quantify human distraction, and apply a non-stationary Markov Decision Process (MDP) model to safely plan UAS motion in the presence of uncertainty. Given the current and next UAS waypoints, assigned by the MDP planner, the paper applies Potryagin’s minimal principle to plan the desired trajectory of the UAS and uses a feedback linearaztion trajectory control to enable UAS with stable tracking of the desired trajectory. Note to Practitioners —Despite advances in aerial robotics, there are still significant barriers for their integration and application into human-centered jobs. Safety-related concerns, potential hazards to labor, and limited mission duration are some major barriers for using such a helpful technology. The main goal of this paper is to come up with a reliable solution for long-term collaboration between humans and unmanned aerial system (UAS) for safe and efficient accomplishment of a human-centered work. To this end, we propose to use Petri Nets to abstractly model and specify human-robot collaboration and effectively plan the tasks assigned to UAS and human co-workers. In this context, human co-workers can do some tasks that are difficult and somehow unsafe to be carried out by UAS. On the other hand, UAS can be deployed to monitor the work environment, recognize human wellness, or carry some small payloads. UAS co-workers enabled with on-board perception capabilities can also predict human coworker intention and quantify human co-worker distraction in real time through online processing of in-situ observations. Furthermore, UAS can safely plan its motion without human intervention or supervision, by combining Markov Decision Process, optimal control, and trajectory tracking control models.


  • A Fixed Air Corridor Model for UAS Traffic Management in Urban Areas

    by Aeris El Asslouj, Harshvardhan Uppaluru, Mohamadreza Ramezani, Ella Atkins, Hossein Rastgoftar

    This paper develops a high-density air corridor traffic flow model for Uncrewed Aircraft System (UAS) operation in urban low altitude airspace. To maximize throughput with safe separation guarantees, we define an airspace spatiotemporal planning problem. We propose a multi-floor UAS coordination structure divided into a finite number of air corridors safely wrapping buildings and obstacles. We use Geographical Information System Lidar data to map buildings and in turn generate air corridors by modeling UAS coordination as ideal fluid flow with streamlines obtained by solving the Laplace partial differential equation. Proper boundary conditions for the differential equations are imposed to direct air corridors along each floor's desired motion direction. For temporal planning, we use four-dimensional path-finding through the corridor network by applying the A* path-finding algorithm through space-time to maximize throughput given each UAS' initial and destination waypoint pair.


  • Open Access

    Safe Human-UAS Collaboration Abstraction

    by Hossein Rastgoftar

    This paper studies the problem of safe humanuncrewed aerial system (UAS) collaboration in a shared work environment. By considering human and UAS as co-workers, we use Petri Nets to abstractly model evolution of shared tasks assigned to human and UAS co-workers. Particularly, the Petri Nets places represent work stations; therefore, the Petri Nets transitions can formally specify displacements between the work stations. The first objective is to incorporate uncertainty regarding the intentions of human co-workers into motion planning for UAS, when UAS co-workers closely interact with human co-workers. To this end, the proposed Petri Nets model uses conflict constructs to represent situations at which UAS deals with incomplete knowledge about human co-worker intention. The second objective is then to plan the motion of the UAS in a resilient and safe manner, in the presence of non-cooperative human co-workers. In order to achieve this objective, UAS equipped with onboard perception and decision-making capabilities are able to, through real-time processing of in-situ observation, predict human intention, quantify human distraction, and apply a non-stationary Markov Decision model to safely plan UAS motion in the presence of uncertainty.