Simone Tolomei

Simone Tolomei

Auto-generated profile, based on publicly available information

Simone Tolomei

Università di Pisa

Research Focus & Metrics

Academic Career Length

3Years (2024-Present)

(Calculated once Scilit publication data is synced: current year − year of first publication + 1)

Representative Publications

View All Publications
5 May 2025 in IEEE Open Journal of the Industrial Electronics Society
Verified
Journal Article
Open access
ASAntonello Scaldaferri
Antonello Scaldaferri
STSimone Tolomei
Simone Tolomei
FIFrancesco Iotti
Francesco Iotti
PGPaolo Gambino
Paolo Gambino
MPMichele Pierallini
Michele Pierallini
MGManolo Garabini
Manolo Garabini
This paper presents the mechanical design of Otto, a lightweight 8-degrees-of-freedom (8-DoF) quadrupedal robot employing Series Elastic Actuators (SEA), and a training framework for learning locomotion control policies in simulation using Reinforcement Learning (RL). Otto's design differs from typical 12-DoF quadrupeds by lacking hip adduction-abduction degrees of freedom. This reduces the robot's cost and weight and increases complexity for tasks involving base rotation and angular twist following. The elastic elements at the joints improve compliance, energy efficiency, safety, and stability, increase robustness, and reduce damage to robot hardware components. Our locomotion control approach leverages RL to optimize policies in simulation, allowing stable and efficient movement despite mechanical constraints, i.e., an 8-DoF quadrupedal robot. Through extensive simulation training, leveraging highly parallel GPU-accelerated robotic simulators, we ensure the policy is well-suited for deployment in real-world scenarios, where accurate motion control is critical for performance. The trained policy is then transferred to the physical robot platform. We demonstrate its effectiveness in various tasks and real-life scenarios with varying payloads and terrains, and compare it with a state-of-the-art model-based method. The results show that Otto, equipped with our RL-based locomotion control, achieves robust performance, compensating for the reality gap and managing the reduced DoF available in Otto.

13 November 2024 in IEEE Robotics and Automation Letters
Verified
Journal Article
Open access
MKMarek Kopicki
Marek Kopicki
SASainul Islam Ansary
Sainul Islam Ansary
STSimone Tolomei
Simone Tolomei
FAFranco Angelini
Franco Angelini
MGManolo Garabini
Manolo Garabini
PSPiotr Skrzypczyński
Piotr Skrzypczyński
We introduce a novel reconfigurable passive joint (RP-joint), which has been implemented and tested on an underactuated three-finger robotic gripper. RP-joint has no actuation, but instead it is lightweight and compact. It can be easily reconfigured by applying external forces and locked to perform complex dexterous manipulation tasks, but only after tension is applied to the connected tendon. Additionally, we present an approach that allows learning dexterous grasps from single examples with underactuated grippers and automatically configures the RP-joints for dexterous manipulation. This is enhanced by integrating kinaesthetic contact optimization, which improves grasp performance even further. The proposed RP-joint gripper and grasp planner have been tested on over 370 grasps executed on 42 IKEA objects and on the YCB object dataset, achieving grasping success rates of 80% and 87%, on IKEA and YCB, respectively.

24 July 2025 in IEEE Robotics & Automation Magazine
Verified
Journal Article
Closed access
STSimone Tolomei
Simone Tolomei
GLGiovanni Di Lorenzo
Giovanni Di Lorenzo
FAFranco Angelini
Franco Angelini
LSLeopoldo de Simone
Leopoldo de Simone
EFE. Fanfarillo
E. Fanfarillo
MGManolo Garabini
Manolo Garabini
This paper presents a novel approach to forest habitat monitoring using robotics and advanced data analysis techniques. We introduce a quadrupedal robot with LiDAR and onboard cameras to collect detailed data about forest structure and composition. The data is then processed using a combination of data analysis techniques and machine learning algorithms to perform a comprehensive dendrometric and floristic survey. Our approach provides an efficient and accurate method for assessing the ecological health of forest ecosystems. This work contributes to the ongoing efforts in habitat conservation and offers a promising tool for future environmental monitoring tasks.

8 May 2025 in IEEE Transactions on Systems, Man, and Cybernetics: Systems
Verified
Journal Article
Closed access
STSimone Tolomei
Simone Tolomei
DBDominik Belter
Dominik Belter
JBJakub Bednarek
Jakub Bednarek
FAFranco Angelini
Franco Angelini
MGManolo Garabini
Manolo Garabini
Mastering rough terrain locomotion is a tough challenge for robots due to its dynamic, unpredictable nature and frequent physical contact. Traditionally, robots rely on carefully planned foot placements to maintain grip and stability. Recent advancements in quadruped robot feet offer diverse shapes and high grip for various terrains. However, control systems and planners often struggle to leverage these varied capabilities, relying instead on simplified foot models e.g., ball-like, flat. The simplified feet models committed to the single shape of the foot can not be used on robots equipped with diverse feet or modern adaptive feet. This work proposes a novel foothold optimization method that efficiently searches for optimal contact points for different foot shapes using a polynomial approximation. The system leverages a Convolutional Neural Network (CNN) trained on simulated data to predict a cost for each candidate foothold. We show that a single neural network can work with different and new foot mechanical designs without retraining the system. We experimentally validate our system on the ANYmal robot using both ball feet and adaptive soft feet, in indoor and outdoor environments, finding that our system improves stability, in terms of pitch and roll angles of the base, with respect to a state-of-the-art method.

5 December 2025 in IEEE Robotics & Automation Magazine
Verified
Journal Article
Closed access
David Rodriguez-CiancaDavid Rodriguez-Cianca
David Rodriguez-Cianca
PSPaulino Sendin
Paulino Sendin
GAGiuseppe Alfonso
Giuseppe Alfonso
DBDavide De Benedittis
Davide De Benedittis
STSimone Tolomei
Simone Tolomei
DTDiego Torricelli
Diego Torricelli
Demonstrating robust and reliable locomotion in real-world environments is essential for validating the performance and safety of legged robots. However, the absence of standardized evaluation frameworks makes it difficult to objectively assess and compare robotic systems under realistic conditions. To address this gap, we propose a comprehensive benchmarking methodology specifically designed for evaluating legged locomotion over irregular and sloped terrains. The framework comprises a modular and easily replicable testbed, a standardized experimental protocol, and a set of performance indicators (PIs) that capture key locomotion metrics such as stability, efficiency, and trajectory accuracy. To demonstrate the utility and applicability of the proposed methodology, we present a case study using the quadruped robot ANYmal-C. We benchmarked its locomotion performance across three different ground materials—flat wood, artificial grass, and TerraSensa—at slope inclinations ranging from 0° to 15°, under two controller configurations (Trekker and Dynamic Gait) and two commanded speeds (0.3 m/s and 0.8 m/s). The results consistently highlight the superior performance of the Trekker controller across all test conditions. Additionally, ground inclination emerged as a more influential factor than surface material, significantly affecting parameters such as slippage and cost of transport (CoT). Higher commanded speeds were also found to enhance locomotion performance under all evaluated scenarios. This work represents a first step toward establishing reproducible and systematic benchmarking practices for legged robotics, offering a practical and scalable approach to performance evaluation in controlled yet realistic environments.

16 March 2026 in IEEE Transactions on Automation Science and Engineering
Verified
Journal Article
Open access
GGGabriele Gambino
Gabriele Gambino
STSimone Tolomei
Simone Tolomei
FAFranco Angelini
Franco Angelini
MGManolo Garabini
Manolo Garabini
Grasping unknown objects in clutter remains challenging due to partial occlusions, self-occlusions, and frequent object interactions during execution. In this paper we present Grasp It Like a Pro 3.0 (GILP 3.0), a lightweight learning-from-demonstration pipeline for closed-loop clutter clearing with unknown objects. The method segments the scene from raw RGB-D point clouds, approximates candidate regions via MVBB decomposition, and predicts grasp pose and interaction wrench from compact MVBB descriptors using two histogram-based gradient-boosted decision-tree regressors trained from a limited number of human demonstrations. Grasp candidates are ranked with execution-oriented feasibility and collision-aware scoring, and the pipeline iteratively reacquires and replans after each attempt to handle object motion and occlusions in clutter. Real-robot experiments on a Franka Emika Panda with Franka Hand in 12 cluttered scenes (47 objects) achieve 95.7% perobject success and a 91.7% scene clearing rate, demonstrating a replicable and data-efficient solution for grasping unknown everyday objects in cluttered environments.