The Use of Open Source Scripting to Reduce Autonomous System Complexity and Cost

Andrea Rowe, United States and Chad Collett, SubC Imaging, Clarenville, NF, Canada
Abstract:
Scripting via an application programming interface (API) allows inspectors to acquire underwater video and images independently from a remote control system by autonomously controlling camera functions. In order to accomplish this, a camera with an embedded CPU and GPU hosts and executes the scripts, running separately from any other system by managing complex workflows, analyzing results, and reacting accordingly.

Traditional camera systems require multiple components and video distribution networks to gather underwater data. Before now, there have been no underwater-specific embedded imaging device products available on the market to replace this process. Scripting allows the camera to operate autonomously without requiring an additional computer to send commands to and conduct scheduling. Because of this, a separate shoreside computer that is running Python, or similar scripts, is not required in order to communicate with the camera via its API. By integrating a scripting feature into a software-defined camera, the sensor payload is greatly reduced.

For research purposes, and in preparation for working with the University of Washington on the Ocean Observatories Initiative, multiple scripting tests were conducted. A time-lapse script, written to take photos and record 4K video, was tested on a parallel setup for one week. The scripts were then updated on the University of Washington’s observatory camera setup where the same results were produced.

Along with ocean observatories, there are many other applications where scripting is beneficial. An example of such an application would be a month-long stop-motion study of subsea geology where a script could be created to put the camera in hibernate mode in order to enhance long durations and battery operation.

In conclusion, with the complexity of multiple systems and high cost removed, allowing for unique and flexible autonomous vision applications, an increase in underwater research and discoveries is highly probable.