Autonomic computing technology for autonomous marine vehicles.

Abstract: Autonomous Marine Vehicles (AMVs) are not only being required to carry out more complex tasks but also longer missions. This mainly requires effective resilient operation and efficient resource management to succeed in persistent presence at sea or ocean with minimal human interaction while maintaining seakeeping performance. Even though some of the current AMVs have a large degree of self-governance, most of them fail to support self-management (e.g. auto-maintenance during pre/in/post-mission phases). Autonomic Computing (AC) basically provides the following self-managing capabilities: self-healing, self-protecting, self-optimizing, and self-configuring. In addition, it provides systems with self-aware, self-adjusted, and self-situated abilities. AC comes from a biological metaphor based on the self-regulating capabilities of the autonomic nervous system in the human body. This paper introduces the AC concept to control architectures of AMVs to endow them with resilience and environmental efficiency. The above capabilities are to help persistent autonomy and automation endure over complex and long AMV operations. This paper presents the architectural aspects, and details of design and realization of this promising AC-based approach. It also discusses four key aspects from existing methodologies and technologies that are potential approaches to support the autonomic control architecture proposed for AMVs. Finally, future research directions are presented. [Copyright &y& Elsevier]