Vincenzo Lippiello

Antoonio Buo, Vittorio Cammarota, Michele Avagnale et al.

In the literature, actor-critic model predictive control (AC-MPC) integrates MPC with reinforcement learning to enable high-performance control of complex dynamical systems. However, its differentiable MPC layer requires repeatedly solving an optimization problem in both the forward and backward passes, leading to substantial training and inference latency. This paper tackles this bottleneck introducing a CUDA-accelerated variant that significantly reduces end-to-end execution time while preserving the control performance of the baseline formulation. Simulation results on an agile drone racing task show that our approach achieves state-of-the-art lap times and near-limit dynamic behaviour with markedly reduced training and inference time.

Andrea Berra, Viswa Narayanan Sankaranarayanan, Achilleas Santi Seisa et al.

The paper introduces a novel framework for safe and autonomous aerial physical interaction in industrial settings. It comprises two main components: a neural network-based target detection system enhanced with edge computing for reduced onboard computational load, and a control barrier function (CBF)-based controller for safe and precise maneuvering. The target detection system is trained on a dataset under challenging visual conditions and evaluated for accuracy across various unseen data with changing lighting conditions. Depth features are utilized for target pose estimation, with the entire detection framework offloaded into low-latency edge computing. The CBF-based controller enables the UAV to converge safely to the target for precise contact. Simulated evaluations of both the controller and target detection are presented, alongside an analysis of real-world detection performance.

Jonathan Cacace, Alberto Finzi, Vincenzo Lippiello

We present a multimodal interaction framework suitable for a human rescuer that operates in proximity with a set of co-located drones during search missions. This work is framed in the context of the SHERPA project whose goal is to develop a mixed ground and aerial robotic platform to support search and rescue activities in a real-world alpine scenario. Differently from typical human-drone interaction settings, here the operator is not fully dedicated to the drones, but involved in search and rescue tasks, hence only able to provide sparse, incomplete, although high-value, instructions to the robots. This operative scenario requires a human-interaction framework that supports multimodal communication along with an effective and natural mixed-initiative interaction between the human and the robots. In this work, we illustrate the domain and the proposed multimodal interaction framework discussing the system at work in a simulated case study.