Max Robotics — Global Service Robot Directory + US Certification

About

Elia Trevisan is a robotics researcher specializing in motion planning, sampling-based model predictive control, and autonomous robot navigation in complex, dynamic environments. His work centers on advancing Model Predictive Path Integral (MPPI) control, a powerful framework for real-time robot decision-making under uncertainty. Trevisan has made significant contributions by demonstrating how GPU-parallelizable physics simulators, such as IsaacGym, can serve as dynamic models within MPPI, enabling more accurate and computationally efficient planning — a line of work that has accumulated over 23 citations across two related publications. His highly cited "Biased-MPPI" paper (25 citations) introduced a novel approach to informing sampling-based controllers by integrating ancillary controllers, substantially improving performance in uncertain, interaction-rich settings. Beyond single-robot navigation, Trevisan has extended MPPI into reactive Task and Motion Planning by combining it with Active Inference, addressing execution robustness in manipulation tasks. His research also tackles the challenge of safely deploying mobile robots among humans, incorporating interaction-awareness and dynamic risk assessment into the planning pipeline. Collectively, his work bridges theoretical advances in stochastic optimal control with practical robotics applications, making him a notable emerging voice in the autonomous systems research community.