Harshvardhan Uppaluru

I am a Postdoctoral Fellow in the Intelligent Multi-level Power-Aware Circuits and Systems (IMPACT) Lab at University of South Alabama, advised by Prof. Jinhui Wang and Prof. Na Gong. IMPACT Lab is associated with Electrical & Computer Engineering Department and College of Engineering.

Previously, I completed my Ph.D. in Aerospace Engineering with a minor in Systems and Industrial Engineering from University of Arizona, where I was advised by Prof. Hossein Rastgoftar. Prior to pursuing Ph.D., I completed my Masters in Robotics Engineering from University of Maryland - College Park. I have a Bachelors degree in Mechatronics Engineering from SRM University.

Apart from research, I play cricket and watch soccer on weekends. Arsenal is my favorite soccer team. In addition, I referee cricket matches. I completed four levels of cricket umpiring: USACUA Preliminary, USACUA Final, WICUA Final, and WICUA Oral & Practical. I have umpired in various leagues such as Minor League Cricket (MiLC), Washington Cricket League (WCL), Arizona Cricket Association (ACA) and Washington Youth Cricket League (WYCL).

Email / GitHub / Google Scholar / LinkedIn / Twitter

Publications

	Drones Practicing Mechanics Harshvardhan Uppaluru and Hossein Rastgoftar International Conference on Unmanned Aircraft Systems (ICUAS), 2023 Paper / Introduced a novel approach for visualizing linear deformation and strain analysis in 2D and 3D motion spaces using quadrotors as particles of a deformable body, providing a practical method for teaching mechanics concepts and ensuring inter-agent collision avoidance through lower bounds on axial strains in continuum deformation coordination.
	Multi-Layer Continuum Deformation Optimization of Multi-Agent Systems Harshvardhan Uppaluru, and Hossein Rastgoftar International Federation of Automatic Control (IFAC), 2023 Paper / Developed a continuum mechanics-based leader-follower paradigm for optimal large-scale multi-agent system (MAS) deformation using quadratic programming, ensuring collision avoidance and minimal deviation from desired trajectory, as demonstrated with 67 quadcopters tracking a helix trajectory in 3D motion space.
	Deep Continuum Deformation Coordination and Optimization with Safety Guarantees Harshvardhan Uppaluru, and Hossein Rastgoftar American Control Conference (ACC), 2023 Paper / Introduced a neural network-based technique for the safe coordination and optimization of a team of N quadcopters that track a known target trajectory in 3D motion space. The algorithm optimizes the weights and biases based on the desired trajectory to track.
	Resilient multi-UAS coordination using cooperative localization Harshvardhan Uppaluru, Hamid Emadi and Hossein Rastgoftar Aerospace Science and Technology (AST), 2022 Paper / A fault-resilient multi-UAS coordination approach using continuum mechanics principles to ensure safety and manage transitions between Homogeneous Deformation Mode and Fault Resilient Mode. The approach utilizes Cooperative Localization for detecting failed UAS and localizing healthy ones at a low computational cost, and has been validated through simulations with various failure scenarios.
	Fast and Safe Aerial Payload Transport in Urban Areas Aeris El Asslouj, Harshvardhan Uppaluru, Hossein Rastgoftar Arxiv Preprint Paper /
	Quadrotor Formation Flying Resilient to Abrupt Vehicle Failures via a Fluid Flow Navigation Function Matthew Romano, Harshvardhan Uppaluru, Hossein Rastgoftar, Ella Atkins Arxiv Preprint Paper / Development and experimental evaluation of a navigation function for quadrotor formation flight resilient to abrupt quadrotor failures and other obstacles.
	A Continuous-Time Optimal Control Approach to Congestion Control Harshvardhan Uppaluru, Hamid Emadi, and Hossein Rastgoftar European Control Conference (ECC), 2022 Paper / A continuous-time optimal control approach for modeling and controlling traffic congestion in a network of interconnected roads using the street map of Phoenix City. The proposed model utilizes mass-flow conservation law, stochastic processes, and boundary control, with constraints ensuring feasibility and traffic backflow avoidance. The approach demonstrates stability in simulation studies, and future work will explore modeling and controlling traffic congestion as a Markov Decision Process.
	A Physics-Based Safety Recovery Approach for Fault-Resilient Multi-Quadcopter Coordination Hamid Emadi, Harshvardhan Uppaluru, and Hossein Rastgoftar American Control Conference (ACC), 2022 Paper / A physics-based approach to deal with abrupt quadrotor failure. It consists of 2 layers: (i) high-level guidance to plan recovery trajectory for healthy quadrotors, (ii) low-level trajectory control design to safely recover as quickly as possible.
	A Physics-Based Data-Driven Approach for Finite Time Estimation of Pandemic Growth Harshvardhan Uppaluru, and Hossein Rastgoftar Modeling, Estimation and Control Conference (MECC), 2022 Paper / A novel physics-based data-driven approach for modeling and predicting the spread of the COVID-19 pandemic in the United States, utilizing spatial and temporal conservation laws. The prediction model demonstrated high accuracy, with relative estimation errors below 1\% for the number of cases, deaths, and recoveries, between March 12, 2020 and October 1, 2021. Furthermore, our results aligned with the observed growth of the pandemic, considering the impact of the delta variant and the emergence of new waves of infections.
	Collision-Free Continuum Deformation Coordination of a Multi-Quadcopter System Using Cooperative Localization Hamid Emadi, Harshvardhan Uppaluru, and Hossein Rastgoftar European Control Conference (ECC), 2022 Paper / We integrate cooperative localization with continuum deformation coordination of a multi-quadrotor system in presence of position uncertainty, to assure safety and optimality of the quadrotor team coordination.