Adapting open-source drone autopilots for real-time iceberg observations

Area of Science:

• Oceanography
• Glaciology
• Robotics

Background:

• Traditional iceberg trackers have limitations in short-term, high-frequency monitoring.
• Commercial trackers often record data over long intervals, unsuitable for detailed drift analysis.

Purpose of the Study:

• To evaluate the effectiveness of drone autopilots for real-time iceberg tracking.
• To demonstrate a cost-efficient method for monitoring iceberg drift and orientation.

Main Methods:

• Utilized an ArduPilot Mega (APM) 2.6 autopilot powered by a lithium-ion battery, housed in a waterproof case.
• Deployed the system onto an iceberg by hand from a boat for real-time data transmission to a ground station.
• Collected conductivity/temperature/depth (CTD) profiles and used APM position data to determine proximity to the iceberg's meltwater plume.

Main Results:

• The APM system provided high-frequency (10-second interval) position data, significantly reducing positional uncertainty compared to traditional methods.
• The system operated continuously for over 10 hours, demonstrating reliability for short-term monitoring.
• Real-time data facilitated safer CTD profiling and accurate calculation of distances to the iceberg.

Conclusions:

• Drone autopilots, like the APM, are a viable, low-cost ($90) tool for real-time iceberg tracking and monitoring iceberg-ocean interactions.
• The described methods are adaptable to all Ardupilot-compatible systems, offering broad applicability.
• This technology enhances safety and provides valuable data for operational iceberg management.