Metric assessment of autonomous capabilities in unmanned maritime vehicles

Insaurralde, Carlos C.; Lane, David L.

Abstract

Autonomous capabilities in Unmanned Maritime Vehicles (UMVs) have been pushing the technology beyond imaginable limits in order to deal with more and more complex ocean and sea missions. However, there is still no agreement on the aspects that should be evaluated to know whether a UMV satisfies the degree or level of autonomy expected. This paper reviews current approaches to assess autonomous behavior in UMVs. It also recaps the different development and operation aspects taken into account by researchers and practitioners to evaluate autonomy in unmanned systems from diverse domains. This provides a baseline to study methodologies to assess maritime autonomy so that metrics and assessment criteria can be defined for UMVs. This paper also presents statistical results from a comparison of existing assessment frameworks for autonomy in different domains. Outcomes from this study are critical to devise guidelines for metric autonomy assessment in UMVs. In addition, a discussion on relevant aspects of a autonomy assessment process for UMVs, and future research directions are presented.

Citation

Insaurralde, C. C., & Lane, D. L. (2014). Metric assessment of autonomous capabilities in unmanned maritime vehicles. Engineering Applications of Artificial Intelligence, 30, 41-48. https://doi.org/10.1016/j.engappai.2013.09.003

• Journal Article Type: Article
• Acceptance Date: Sep 8, 2013
• Online Publication Date: Oct 8, 2013
• Publication Date: Apr 1, 2014
• Deposit Date: Jan 29, 2020
• Journal: Engineering Applications of Artificial Intelligence
• Print ISSN: 0952-1976
• Publisher: Elsevier
• Peer Reviewed: Peer Reviewed
• Volume: 30
• Pages: 41-48
• DOI: https://doi.org/10.1016/j.engappai.2013.09.003
• Keywords: Maritime autonomy, Unmanned marine vehicles, Autonomy metrics, Autonomous capabilities, Autonomy assessment criteria
• Public URL: https://uwe-repository.worktribe.com/output/5263666
• Publisher URL: https://www.sciencedirect.com/journal/engineering-applications-of-artificial-intelligence