Plenary Lectures

Distributed formation control is a fundamental functionality of multi-robot systems, enabling autonomous vehicles to maneuver cohesively without relying on centralized infrastructure. However, achieving robust performance in the presence of relative measurement mismatches, parametric uncertainties, and heterogeneous sensor topologies, while safely navigating obstacle-strewn environments, remains a major barrier to real-world adoption. This plenary talk presents recent theoretical and practical advances aimed at closing these critical gaps.

First, we introduce adaptive and dynamical control laws designed to eliminate measurement mismatches, compensate for parametric uncertainties, and reject external disturbances. We then delve into the challenges of operating with heterogeneous sensor systems and present approaches to mitigate these problems. Extending this work to use a low-cost sensing solution, we examine the use of vision-based relative information to achieve the formation. To guarantee safe transit, we discuss the integration of distributed Control Barrier Functions directly into the formation control design. Finally, we introduce a paradigm-shifting formation control problem by moving beyond rigid geometric shapes to focus on how a swarm can dynamically achieve a desired spatial distribution function across a given space.

The merging of artificial intelligence (AI) and control systems could generally arise into two different forms : Data driven (AI based) control and integration of AI and control, particularly in robotics and autonomous physical systems, which transforms machines from rigid, pre-programmed tools into intelligent agents. In the former data-driven (AI-based) control, various methods were developed to replace or augment traditional physics-based models by learning controller actions directly from measured system data. It leverages machine learning—particularly Reinforcement Learning (RL) and neural networks—to optimize, adapt, and control dynamic systems without needing exact mathematical equations.

In this talk, I will focus on the latter approaches where AI enables perception, reasoning, and high-level decision-making, while traditional control theory ensures physical stability, precision, and safe trajectory execution. Modern autonomous systems operate on a unified framework where data-driven learning complements mathematical models: 1) Perception & reasoning : using machine learning and computer vision, AI processes raw sensory data to identify objects, classify environments, and predict dynamic changes in the surroundings, 2) Planning: AI algorithms (such as reinforcement learning) determine the optimal high-level strategy, like determining the most efficient path around obstacles in real-time. 3) Execution & stability : mathematical control systems such as LQG, H,  controllers or MPC calculate the exact physical forces, motor torques, or hydraulic pressures required to execute the AI's commands smoothly. Key benefits of integration of AI and control includes : 1) adaptability: robots can handle unstructured, dynamic environments. Instead of following a rigid path, a robotic arm can adjust its grip if an object slips, or a drone can navigate a forest without a pre-mapped route; 2) safety guarantees: by coupling neural networks with mathematical "barrier certificates" (control algorithms that prevent a system from crossing safety limits), developers can ensure the robot acts intelligently without violating physical constraints.

The integration of AI and control can be found in a number of use cases such as autonomous vehicles, industrial automation (cobots), aerial, ground and marine drones. In autonomous vehicles, AI handles surrounding object detection, classification semantic segmentation and tracking, as well as route planning, while the control system manages the precise steering angle, braking, and throttle. In industrial automation AI allows robotic arms to visually adapt to assembly lines, working safely alongside human workers without needing physical safety cages. In aerial, ground and marine drones, AI handles mapping and target tracking, while advanced control loops correct for wind disturbances, turbulence, uneven terrains or sea currents.

I will also share my recent works on autonomous vehicle, underwater robot, and collaboration among ground and aerial robots where integration of AI and control leads to adaptive decision making and intelligent behaviour in unstructured and dynamic environments.

Road infrastructure remains significantly underutilized in current traffic systems. Mixed-autonomy traffic, where connected and automated vehicles (CAVs) interact with human-driven vehicles, offers a unique opportunity to improve system-level efficiency, resilience, and safety. In this context, CAVs can be leveraged not only as transportation agents but also as distributed sensors and actuators, enabling real-time state estimation and feedback control of large-scale traffic networks. This presentation introduces a physics-informed learning and control framework for mixed-autonomy traffic systems. By embedding traffic flow dynamics into machine learning models, we develop structured, data-efficient representations that generalize across operating conditions while preserving physical consistency. These models enable the design of scalable feedback control strategies that proactively mitigate congestion and respond to disturbances under uncertainty. We provide a comparative analysis of model architectures from a control perspective, highlighting trade-offs in expressiveness, interpretability, and robustness. Furthermore, we address the challenges of human–automation interaction by integrating formal safety guarantees into the control design and by incorporating teleoperation as a supervisory fallback mechanism in safety-critical scenarios. The proposed framework is validated through extensive real-world demonstrations conducted in collaboration with Swedish industry partners, illustrating its potential for deployment in next-generation intelligent transportation systems.

This talk will provide an overview on the recent research on multi-agent systems operating in hostile environments. In the context of consensus problems, we will focus on the influence of misbehaving agents capable to inject false data in their transmissions and how to mitigate such cyber attacks by the approach of the so-called mean subsequence reduced algorithms and their variants. Agents equipped with such algorithms will ignore their neighbors taking outlying state values. We will see that characterizations on the properties necessary for network topologies can be established, and moreover that network resiliency can be enhanced when more communication and computational resources are available. This approach originates in the area of distributed algorithms in computer science, but recent studies in systems control have brought notable advances. We will further discuss extensions of such algorithms to problems of averaging, parameter estimation, and clock synchronization in wireless sensor networks.

Bruno Siciliano is professor of robotics and control at the University of Naples Federico II. He is also Honorary Professor at the University of Óbuda where he holds the Kálmán Chair. His research interests in robotics include manipulation and control, human–robot cooperation, and service robotics. Fellow of the scientific societies IEEE, ASME, IFAC, AAIA, AIIA, NAAI, he received numerous international prizes and awards, including the recent 2024 Pioneer Award in Robotics and Automation. He was President of the IEEE Robotics and Automation Society from 2008 to 2009. He has delivered more than 150 keynotes and has published more than 350 papers and 27 books. His book “Robotics” is among the most adopted academic texts worldwide, while his edited volume “Springer Handbook of Robotics” received the highest recognition for scientific publishing: the 2008 PROSE Award for Excellence in Physical Sciences & Mathematics. His team has received more than 25 million Euro funding in the last 18 years from competitive European research projects, including an Advanced Grant and a Synergy Grant from ERC. More details are available at http://wpage.unina.it/sicilian/

This talk presents research results @ PRISMA Lab on robot manipulation and control. The talk is organized in four parts. In the first part, control techniques for dynamic nonprehensile manipulation are presented. The second part of the talk focuses on how to merge learning and model-based strategies to provide autonomy to robot manipulation. In the third part, several aerial robotics applications for inspection and maintenance are surveyed. The fourth part of the talk deals with recent advances on shared control including haptic guidance.

Masayoshi Tomizuka received his B.S. and M.S. from Keio University in 1968 and 1970, respectively. He received his Ph. D. from MIT in 1974, after which he joined the ME Department at UC Berkeley. Here, he served as the Vice Chair of Instruction from Dec. 1989 to Dec. 1991, and as the Vice Chair of graduate studies from Jul. 1995 to Dec. 1996.

He is currently the Associate Dean of Academic Affairs for the College of Engineering at UC Berkeley. From 2009 to 2011, he was the Executive Associate Dean for the College of Engineering at UC Berkeley. He also served as Program Director of the Dynamic Systems and Control Program at the National Science Foundation from Sept. 2002 to Dec. 2004.

Prof. Tomizuka’s research interests include optimal and adaptive control, digital control, signal processing, motion control, mechatronics and their applications in robotics, manufacturing, data storage devices, vehicles, and human-machine systems.

In this talk, I will review my career path that started at Keio University. I chose my current specialization of control theory. The path continued through the master of science study at Keio, PhD study at the Massachusetts Institute of Technology (MIT) and teaching at the University of California, Berkeley (UCB). At UCB, I founded and directed the Mechanical Systems Control (MSC) Laboratory where research on mechatronics is conducted. The current research program at the MSC Laboratory is focused on two topics: intelligent robot systems and autonomous driving systems. I have had the privilege of interacting with young researchers and students. By now, I have supervised the research of 155 PhD students to completion. The talk includes take away and suggestions to young researchers.