Don't Make the Same Mistakes Again and Again: Learning Local Recovery Policies for Navigation from Human Demonstrations

Del Duchetto, Francesco and Kucukyilmaz, Ayse and Iocchi, Luca and Hanheide, Marc (2018) Don't Make the Same Mistakes Again and Again: Learning Local Recovery Policies for Navigation from Human Demonstrations. IEEE Robotics and Automation Letters . ISSN 2377-3774

Full content URL: https://ieeexplore.ieee.org/document/8423079/

DocumentsAudio/Video Files
Don't Make the Same Mistakes Again and Again: Learning Local Recovery Policies for Navigation from Human Demonstrations

Request a copy
[img] PDF
DelDucetto-RAL18-GPrecovery-preprint.pdf - Whole Document
Restricted to Repository staff only until 31 July 2019.

930kB
[img] Video (MP4)
RAL_final_marrtino_2.mp4 - Other
Restricted to Repository staff only until 31 July 2019.

11MB
Item Type:Article
Item Status:Live Archive

Abstract

In this paper, we present a human-in-the-loop learning framework for mobile robots to generate effective local policies in order to recover from navigation failures in long-term autonomy. We present an analysis of failure and recovery cases derived from long-term autonomous operation of a mobile robot, and propose a two-layer learning framework that allows to detect and recover from such navigation failures. Employing a learning by demonstration (LbD) approach, our framework can incrementally learn to autonomously recover from situations it initially needs humans to help with. The learning framework allows for both real-time failure detection and regression using Gaussian processes (GPs). Our empirical results on two different failure scenarios indicate that given 40 failure state observations, the true positive rate of the failure detection model exceeds 90%, ending with successful recovery actions in more than 90% of all detected cases.

Additional Information:© 2018 IEEE
Keywords:Navigation, Task analysis, Robot sensing systems, Mobile robots, Robustness, Trajectory, Service Robots, Failure Detection and Recovery, Learning from demonstration
Subjects:G Mathematical and Computer Sciences > G760 Machine Learning
H Engineering > H670 Robotics and Cybernetics
H Engineering > H671 Robotics
Divisions:College of Science > School of Computer Science
ID Code:32850
Deposited On:07 Aug 2018 11:05

Repository Staff Only: item control page