19 papers
Submodular Policy Learning for Distributed Task Allocation in Open Multi-Agent Systems
Jing Liu, Luca Ballotta, Yangyang Yang +3
This paper studies policy learning for distributed task allocation in open multi-agent systems, where agents may join and leave in a time-varying fashion, with submodular stage tea…
Hierarchical Topology-Aware Planning and Control of Underwater Vehicle-Manipulator Systems in Confined Environments
Mohamed Abdelwahab, Ruggero Carli, Damiano Varagnolo +1
This paper addresses autonomous intervention with an underwater vehicle--manipulator system (UVMS) in confined, cluttered, and partially known environments, where poor maneuverabil…
Learning Structured Robot Policies from Vision-Language Models via Synthetic Neuro-Symbolic Supervision
Alessandro Adami, Tommaso Tubaldo, Marco Todescato +2
Vision-Language Models (VLMs) have recently demonstrated strong capabilities in mapping multimodal observations to robot behaviors. However, most current approaches rely on end-to-…
Submodular Multi-Agent Policy Learning for Online Distributed Task Allocation in Open Multi-Agent Systems
Jing Liu, Yangyang Yang, Luca Ballotta +3
This paper studies multi-agent reinforcement learning with submodular team utilities for online distributed task allocation. In this setting, each agent selects one action from a l…
Real2Sim via Active Perception with Behavior Trees Automatically Generated by VLMs
Alessandro Adami, Sebastian Zudaire, Ruggero Carli +1
Constructing physically accurate simulation environments (Real2Sim) traditionally relies on manual system identification or rigid, exhaustive exploration routines. These task-agnos…
Learning Stack-of-Tasks Management for Redundant Robots
Alessandro Adami, Aris Synodinos, Matteo Iovino +2
This paper presents a novel framework for automatically learning complete Stack-of-Tasks (SoT) controllers for redundant robotic systems, including task priorities, activation logi…
A Robust Controller based on Gaussian Processes for Robotic Manipulators with Unknown Uncertainty
Giulio Giacomuzzo, Mohamed Abdelwahab, Marco Calì +2
In this paper, we propose a novel learning-based robust feedback linearization strategy to ensure precise trajectory tracking for an important family of Lagrangian systems. We assu…
Finetuning Deep Reinforcement Learning Policies with Evolutionary Strategies for Control of Underactuated Robots
Marco Calì, Alberto Sinigaglia, Niccolò Turcato +2
Deep Reinforcement Learning (RL) has emerged as a powerful method for addressing complex control problems, particularly those involving underactuated robotic systems. However, in s…
Accelerating Model-Based Reinforcement Learning using Non-Linear Trajectory Optimization
Marco Calì, Giulio Giacomuzzo, Ruggero Carli +1
This paper addresses the slow policy optimization convergence of Monte Carlo Probabilistic Inference for Learning Control (MC-PILCO), a state-of-the-art model-based reinforcement l…
Learning global control of underactuated systems with Model-Based Reinforcement Learning
Niccolò Turcato, Marco Calì, Alberto Dalla Libera +3
This short paper describes our proposed solution for the third edition of the "AI Olympics with RealAIGym" competition, held at ICRA 2025. We employed Monte-Carlo Probabilistic Inf…
Towards Autonomous Reinforcement Learning for Real-World Robotic Manipulation with Large Language Models
Niccolò Turcato, Matteo Iovino, Aris Synodinos +3
Recent advancements in Large Language Models (LLMs) and Visual Language Models (VLMs) have significantly impacted robotics, enabling high-level semantic motion planning application…
Distributed Constraint-Coupled Optimization: Harnessing ADMM-consensus for robustness
Mohamed Abdelmouamin Messilem, Guido Carnevale, Ruggero Carli
In this paper, we consider a network of agents that jointly aim to minimise the sum of local functions subject to coupling constraints involving all local variables. To solve this…
Data efficient Robotic Object Throwing with Model-Based Reinforcement Learning
Niccolò Turcato, Giulio Giacomuzzo, Matteo Terreran +3
Pick-and-place (PnP) operations, featuring object grasping and trajectory planning, are fundamental in industrial robotics applications. Despite many advancements in the field, PnP…
Continual Learning for Behavior-based Driver Identification
Mattia Fanan, Davide Dalle Pezze, Emad Efatinasab +3
Behavior-based Driver Identification is an emerging technology that recognizes drivers based on their unique driving behaviors, offering important applications such as vehicle thef…
Edge Delayed Deep Deterministic Policy Gradient: efficient continuous control for edge scenarios
Alberto Sinigaglia, Niccolò Turcato, Ruggero Carli +1
Deep Reinforcement Learning is gaining increasing attention thanks to its capability to learn complex policies in high-dimensional settings. Recent advancements utilize a dual-netw…
AI Olympics challenge with Evolutionary Soft Actor Critic
Marco Calì, Alberto Sinigaglia, Niccolò Turcato +2
In the following report, we describe the solution we propose for the AI Olympics competition held at IROS 2024. Our solution is based on a Model-free Deep Reinforcement Learning ap…
Learning control of underactuated double pendulum with Model-Based Reinforcement Learning
Niccolò Turcato, Alberto Dalla Libera, Giulio Giacomuzzo +2
This report describes our proposed solution for the second AI Olympics competition held at IROS 2024. Our solution is based on a recent Model-Based Reinforcement Learning algorithm…
Adaptive Robust Controller for handling Unknown Uncertainty of Robotic Manipulators
Mohamed Abdelwahab, Giulio Giacomuzzo, Alberto Dalla Libera +1
The ability to achieve precise and smooth trajectory tracking is crucial for ensuring the successful execution of various tasks involving robotic manipulators. State-of-the-art tec…
Exploiting Estimation Bias in Clipped Double Q-Learning for Continous Control Reinforcement Learning Tasks
Niccolò Turcato, Alberto Sinigaglia, Alberto Dalla Libera +2
Continuous control Deep Reinforcement Learning (RL) approaches are known to suffer from estimation biases, leading to suboptimal policies. This paper introduces innovative methods…