MIND-Stack: Modular, Interpretable, End-to-End Differentiability for Autonomous Navigation

TL;DR

MIND-Stack is a modular, end-to-end differentiable framework for autonomous navigation that combines interpretability with learning capabilities, improving control accuracy and enabling real-world deployment on limited hardware.

Contribution

The paper introduces MIND-Stack, a novel modular and differentiable autonomous navigation stack that integrates interpretable states with end-to-end learning, bridging the gap between rule-based and neural approaches.

Findings

Localization reduces downstream control error.

Better performance than state-of-the-art algorithms.

Successful sim-to-real deployment on embedded platform.

Abstract

Developing robust, efficient navigation algorithms is challenging. Rule-based methods offer interpretability and modularity but struggle with learning from large datasets, while end-to-end neural networks excel in learning but lack transparency and modularity. In this paper, we present MIND-Stack, a modular software stack consisting of a localization network and a Stanley Controller with intermediate human interpretable state representations and end-to-end differentiability. Our approach enables the upstream localization module to reduce the downstream control error, extending its role beyond state estimation. Unlike existing research on differentiable algorithms that either lack modules of the autonomous stack to span from sensor input to actuator output or real-world implementation, MIND-Stack offers both capabilities. We conduct experiments that demonstrate the ability of the localization module to reduce the downstream control loss through its end-to-end differentiability while offering better performance than state-of-the-art algorithms. We showcase sim-to-real capabilities by deploying the algorithm on a real-world embedded autonomous platform with limited computation power and demonstrate simultaneous training of both the localization and controller towards one goal. While MIND-Stack shows good results, we discuss the incorporation of additional modules from the autonomous navigation pipeline in the future, promising even greater stability and performance in the next iterations of the framework.