VIP-Loco: A Visually Guided Infinite Horizon Planning Framework for Legged Locomotion

Fuente: arXiv
Saved in:
Bibliographic Details
Main Authors: Shirwatkar, Aditya, Gupta, Satyam, Kolathaya, Shishir
Format: Preprint
Published: 2026
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866910053755256832
author Shirwatkar, Aditya
Gupta, Satyam
Kolathaya, Shishir
author_facet Shirwatkar, Aditya
Gupta, Satyam
Kolathaya, Shishir
contents Perceptive locomotion for legged robots requires anticipating and adapting to complex, dynamic environments. Model Predictive Control (MPC) serves as a strong baseline, providing interpretable motion planning with constraint enforcement, but struggles with high-dimensional perceptual inputs and rapidly changing terrain. In contrast, model-free Reinforcement Learning (RL) adapts well across visually challenging scenarios but lacks planning. To bridge this gap, we propose VIP-Loco, a framework that integrates vision-based scene understanding with RL and planning. During training, an internal model maps proprioceptive states and depth images into compact kinodynamic features used by the RL policy. At deployment, the learned models are used within an infinite-horizon MPC formulation, combining adaptability with structured planning. We validate VIP-Loco in simulation on challenging locomotion tasks, including slopes, stairs, crawling, tilting, gap jumping, and climbing, across three robot morphologies: a quadruped (Unitree Go1), a biped (Cassie), and a wheeled-biped (TronA1-W). Through ablations and comparisons with state-of-the-art methods, we show that VIP-Loco unifies planning and perception, enabling robust, interpretable locomotion in diverse environments.
format Preprint
id arxiv_https___arxiv_org_abs_2603_14345
institution arXiv
publishDate 2026
record_format arxiv
spellingShingle VIP-Loco: A Visually Guided Infinite Horizon Planning Framework for Legged Locomotion
Shirwatkar, Aditya
Gupta, Satyam
Kolathaya, Shishir
Robotics
Perceptive locomotion for legged robots requires anticipating and adapting to complex, dynamic environments. Model Predictive Control (MPC) serves as a strong baseline, providing interpretable motion planning with constraint enforcement, but struggles with high-dimensional perceptual inputs and rapidly changing terrain. In contrast, model-free Reinforcement Learning (RL) adapts well across visually challenging scenarios but lacks planning. To bridge this gap, we propose VIP-Loco, a framework that integrates vision-based scene understanding with RL and planning. During training, an internal model maps proprioceptive states and depth images into compact kinodynamic features used by the RL policy. At deployment, the learned models are used within an infinite-horizon MPC formulation, combining adaptability with structured planning. We validate VIP-Loco in simulation on challenging locomotion tasks, including slopes, stairs, crawling, tilting, gap jumping, and climbing, across three robot morphologies: a quadruped (Unitree Go1), a biped (Cassie), and a wheeled-biped (TronA1-W). Through ablations and comparisons with state-of-the-art methods, we show that VIP-Loco unifies planning and perception, enabling robust, interpretable locomotion in diverse environments.
title VIP-Loco: A Visually Guided Infinite Horizon Planning Framework for Legged Locomotion
topic Robotics
url https://arxiv.org/abs/2603.14345