MIND-Stack: Modular, Interpretable, End-to-End Differentiability for Autonomous Navigation
This addresses the problem of balancing interpretability and learning efficiency in autonomous navigation for robotics, though it is incremental by extending differentiable algorithms to full-stack implementation.
The paper tackles the challenge of developing robust navigation algorithms by introducing MIND-Stack, a modular and interpretable software stack with end-to-end differentiability, which enables the localization module to reduce control error and outperforms state-of-the-art methods in experiments.
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.