Marine Science from an Armchair: A Unified Piloting Framework for Autonomous Marine Vehicles

The UK's National Marine Equipment Pool contains over 40 Autonomous Marine Vehicles (AMVs) for use by the UK marine science community. In order to deliver science consistently, the different types of vehicle each require considerable training and a large pool of trained operators to monitor and pilot these vehicles whilst at sea. However, the number of deployments that can be run in parallel is currently limited by the number of human pilots available and the considerable training overhead.At present, in order to pilot a heterogeneous fleet of AMVs, technical knowledge about each platform and a suite of different software is required, restricting the use of mobile devices. The disparate tools and difference in workflow between vehicle types currently makes it challenging to incorporate external knowledge of the environment within the piloting flow, such as ocean models and remote sensing data, which can be invaluable to the pilots and scientists alike. In this paper, we discuss the existing piloting workflows and present the development of a new unified piloting framework. Our work allows pilots to monitor and control a range of AMV types from any web-capable device. One of the overarching aims of the project is to greatly reduce the barrier of entry for piloting AMVs, allowing scientists and end-users to be more involved in the piloting process, subsequently allowing us to increase the efficiency of deployed vehicles, collect the most purposeful data and make best use of the different capabilities offered by each family of vehicle in multi-AMV operations