ESCAPE: Episodic Spatial Memory and Adaptive Execution Policy for Long-Horizon Mobile Manipulation

TL;DR

ESCAPE introduces a novel framework combining persistent 3D spatial memory and adaptive policies to improve long-horizon mobile manipulation in complex indoor environments, achieving state-of-the-art results.

Contribution

The paper presents ESCAPE, a new approach that integrates episodic spatial memory with an adaptive execution policy for enhanced robustness and flexibility.

Findings

Achieves 65.09% success in seen environments on ALFRED benchmark.

Attains 60.79% success in unseen environments, demonstrating generalization.

Reduces redundant exploration, improving path-length efficiency.

Abstract

Coordinating navigation and manipulation with robust performance is essential for embodied AI in complex indoor environments. However, as tasks extend over long horizons, existing methods often struggle due to catastrophic forgetting, spatial inconsistency, and rigid execution. To address these issues, we propose ESCAPE (Episodic Spatial Memory Coupled with an Adaptive Policy for Execution), operating through a tightly coupled perception-grounding-execution workflow. For robust perception, ESCAPE features a Spatio-Temporal Fusion Mapping module to autoregressively construct a depth-free, persistent 3D spatial memory, alongside a Memory-Driven Target Grounding module for precise interaction mask generation. To achieve flexible action, our Adaptive Execution Policy dynamically orchestrates proactive global navigation and reactive local manipulation to seize opportunistic targets. ESCAPE achieves state-of-the-art performance on the ALFRED benchmark, reaching 65.09% and 60.79% success rates in test seen and unseen environments with step-by-step instructions. By reducing redundant exploration, our ESCAPE attains substantial improvements in path-length-weighted metrics and maintains robust performance (61.24% / 56.04%) even without detailed guidance for long-horizon tasks.