|
RAFFAELE ROMAGNOLI
Research Scientist
Conference Editorial Board Member, IEEE Control Systems Society
RH 242
5000 Forbes Ave
Pittsburgh, PA 15213, USA
e-mail: rromagno (at) andrew.cmu.edu
|
Description: The goal is to develop novel tools that can ensure safety and security for controlling and sensing distributed systems that use AI to improve performance. CPS applications include autonomous vehicles such as UAVs and UGVs.
Topics: Bayesian filtering, stochastic control, machine learning approaches, and neural networks, state estimation, distributed robotics systems, networked control systems, real-time computing, cloud-edge computing.
This research is in collaboration with Prof. Anthony Rowe, Prof. Soummya Kar, and Prof. Yorie Nakahira.
This research is in collaboration with Prof. Anthony Rowe and Bosch company. Edge-cloud computing is revolutionizing many applications such as autonomous vehicles, robotics, industrial manufacturing, etc. This project focuses on controlling networked systems by leveraging edge-cloud computing resources to achieve enhanced monitoring and coordination while taking into account resource constraints such as energy and latency.
This project started as a research project for master students and is also in collaboration with people from the Institute for Human & Machine Cognition (IHMC). Cyber-attacks on run-time control software are a real threat in Cyber-physical systems (CPSs) such as unmanned aerial vehicles (UAVs). One effective solution is the application of software rejuvenation that consists of the periodic software refresh of the run-time control software, which can make the CPS resilient to attacks. Despite its effectiveness against attacks, the overall performance of the system during the mission drastically deteriorates. Deep RL has recently demonstrated its effectiveness in many control applications. In this project, we aim to apply deep RL on drones that implement software rejuvenation against cyber-attacks with the goal to improve the performance of the drone.
This project is in collaboration with Prof. Jason Lohmueller University of Pittsburgh, UPMC Hillman Cancer Center. Chimeric antigen receptor T cells (CAR-T) are genetically modified to recognize tumor-specific antigens and generate a cytotoxic environment for the tumor. The goal is to develop a control theoretic model of the Car-T cells and cell signaling. The model will be used to extract control theoretical properties that can be useful for the development of new kinds of Car-T cells to reduce solid tumors.
This project is in collaboration with the Software Engineering Institute, Carnegie Mellon University. An assurance architecture is a structured set of interdependent analysis from many domains such as model checking, control, real-time theory, security, etc. The goal is to provide the foundations for developing and evolving formally assured systems, thereby enabling rapid deployment of innovative capability to the field. Our goal is how to check whether the assurance architectures are correct, and my role is to define the model problem in the field of control theory and develop connections with Scientific ML for providing tools that can check the correctness of the assurance architecture.
Projects:
Projects:
Projects:
Fall 2022 | I completed the Future Faculty Program, Eberly Center for Teaching Excellence and Educational Innovation, Carnegie Mellon University. |
---|---|
Feb. 2015 | Ph.D., Control System and Automation, Department of Information Engineering, Univestita' Politecnica delle Marche, Ancona, Italy. |
Dissertation: |
|
Oct. 2011 | M.S.,Industrial Automation Engineering, Universita' Politecnica delle Marche, Ancona, Italy. 110/110 with honors |
Nonlinear Control
The course will serve as an introduction to the analysis and design of nonlinear systems and nonlinear control systems. It will provide theoretical and practical tools to deal with nonlinearities which play a crucial role in many disciplines such as engineering, robotics, physics, biology, economy, etc. just to name a few. In this way, students will have the possibility to choose their future in a wide range of different directions. Learning how to apply theoretical knowledge in hands-on projects also helps students to find solutions that can be immediately deployed in a real-world scenario. The course consists of three main parts: a) mathematical foundations and analysis of nonlinear differential equations, b) stability theory of nonlinear systems and c) nonlinear feedback control. Throughout the course, students will implement the theory by using simulation tools and they will work on a real-world project that consists of the design and implementation of a nonlinear control system for UAVs by using the PX4 flight control architecture and jMAVSim simulator. Students will share their experiences and deliver their results in engaging discussions with the rest of the class. Briefly, the main topics are the existence and uniqueness of the solution of Ordinary Differential Equations (ODEs), stability analysis using Lyapunov theory and Invariance theory, design of stabilizing controllers using a variety of methods selected from linearization, sliding modes, feedback linearization and geometric control.
See Google Scholar and Scopus for the full list of papers.
See Google Scholar and Scopus for the full list of papers.