Sensor and Behavior Frameworks for Implementing Backseat Driver on a Slocum Glider

The Slocum glider's long-duration and cost-effective data collection capability, coupled with remote operation from shore, makes them ideal platforms for oceanographic research. However, once submerged in standard operation, instrument operation, waypoints, behaviors, and more are fixed until the vehicle surfaces. This lack of underwater adaptive control can lead to limitations in scientific sampling, limiting the glider's ability to meet the mission's scientific goals. In this work, we present a manual for implementing the Slocum Backseat Driver (BSD) architecture on a G3S Slocum glider. It contains examples for configuring an external controller (a Raspberry Pi 4 Model B), managing file transfers, incorporating real-time sensor data for decision-making, and advancing the ocean simulation capability of a glider simulator using a Slocum Fleet Mission Control (SFMC) script. Four simulations were ran, two utilizing the Slocum BSD architecture to autonomously navigate with a heading orthogonal from the Depth-Averaged Current (DAC) value calculated by the glider. The missions utilizing the BSD architecture updated the heading less frequently and did so with a more optimal average heading. Through using the Slocum BSD architecture, we can fully realize the glider's true capability for long-duration and cost-effective persistent oceanographic sampling presence in challenging regions throughout the globe. This research and accompanying manual provide a starting point for glider pilots, technicians, scientists, and others when working through how to use the Slocum glider's BSD architecture.