Prototyping an Autonomous Underwater Vehicle to Support Port Protection Challenges

Port infrastructure plays an important role in our global economy and should be protected accordingly from both surface and subsurface threats. This paper describes a potential autonomous underwater vehicle (AUV) that was designed and fabricated as a prototype to complete a subset of these port protection functions. To decompose the prototype design into manageable research topics, we broke the problem into three independent, yet related, efforts - perception, autonomy, and platform design. The perception system paired machine learning algorithms with an imaging sonar processing capability to detect and classify subsurface objects in real time. Onboard autonomy functionality, developed through software, included the Autonomy Tool Kit (ATK) and the Mission Oriented Operating Software (MOOS). The platform structure was designed using commercially available acrylic tubing and custom additively manufactured (AM) pressure vessel sections to house all critical sensors and electronics. The design approach is discussed in this paper, as well as fabrication and testing of the prototype AUV. Testing included simulation-based, software in the loop, hardware in the loop, and full integration test events. By providing details about selected hardware and software components used for this prototype development we hope to enable future prototypes in the research community for challenging subsurface problem sets.