Embedded out-of-distribution detection on an autonomous robot platform

TL;DR

This paper presents an unsupervised deep neural network-based out-of-distribution detector implemented on a real-time embedded autonomous robot, demonstrating high success in emergency stopping and addressing resource challenges in ROS middleware.

Contribution

It introduces a novel real-time OOD detection method on an embedded autonomous robot platform with practical evaluation and analysis of resource constraints.

Findings

Achieved 87.5% success rate in emergency stopping.

Demonstrated feasibility of real-time OOD detection on embedded systems.

Analyzed resource challenges specific to ROS middleware.

Abstract

Machine learning (ML) is actively finding its way into modern cyber-physical systems (CPS), many of which are safety-critical real-time systems. It is well known that ML outputs are not reliable when testing data are novel with regards to model training and validation data, i.e., out-of-distribution (OOD) test data. We implement an unsupervised deep neural network-based OOD detector on a real-time embedded autonomous Duckiebot and evaluate detection performance. Our OOD detector produces a success rate of 87.5% for emergency stopping a Duckiebot on a braking test bed we designed. We also provide case analysis on computing resource challenges specific to the Robot Operating System (ROS) middleware on the Duckiebot.