Autonomous and adaptive oceanographic feature tracking on board autonomous underwater vehicles

Abstract

Abstracte capabilities of autonomous underwater vehicles (AUVs) and their ability to perform tasks both autonomously and adaptively are rapidly improving, and the desire to quickly and efficiently sample the ocean environment as Earth's climate changes and natural disasters occur has increased significantly in the last decade. As such, this thesis proposes to develop a method for single and multiple AUVs to collaborate autonomously underwater while autonomously adapting their motion to changes in their local environments, allowing them to sample and track various features of interest with greater efficiency and synopticity than previously possible with preplanned AUV or ship-based surveys. is concept is demonstrated to work in field testing on multiple occasions: with a single AUV autonomously and adaptively tracking the depth range of a thermocline or acousticline, and with two AUVs coordinating their motion to collect a data set in which internal waves could be detected. is research is then taken to the next level by exploring the problem of adaptively and autonomously tracking spatiotemporally dynamic underwater fronts and plumes using individual and autonomously collaborating AUVs. Stephanie entered the Massachusetts Institute of Technology/Woods Hole Oceanographic Institution Joint Graduate Program following college to study Oceanographic Engineering. Her Doctoral research, which comprises this thesis, has focused on using autonomous underwater vehicles (AUVs) to perform autonomous and environmentally adaptive sampling of the ocean environment, focusing on underwater feature detection and tracking for more efficient and synoptic data collection with AUVs. Stephanie has enjoyed being on and near the ocean, working with oceanographic vehicles and instrumentation, and seeing her complex underwater vehicle missions work out in the water, and she has had the great opportunity of participating in over a dozen research cruises since she first started working with underwater vehicles in college. roughout graduate school, Stephanie has continued her many hobbies and activities, adding hiking, farming, sailing, and woodworking to the list. I would also like to thank all the members of the LAMSS lab past and present: ank you Toby for fixing nearly any software problem we threw at you on cruise and in the office, and for completing our A-team for every cruise we were on together. ank you also for being my partner in crime on cruise, on travel, and in life. I could not have completed this research nearly as efficiently without you there to devise all sorts of work-arounds when I wanted extra features to make my virtual experiments easier to run. ank you to Erin and Sheida for being both great friends and lab mates and for teaching me that it is okay to not be the only woman in an engineering lab! anks Stephanie for always being excited about my research and hosting craft nights. om, thanks for being such a quick learner on cruise and generally a fun person to hang out with. AcknowledgmentsYou will be a good rep...

Front delineation and tracking with multiple underwater vehicles

“…Previous work on autonomous front tracking in simulation includes Petillo () and Petillo, Schmidt, Lermusiaux, Yoerger, and Balasuriya () who use single and multiple autonomously coordinated AUVs following a dynamic zigzag across a threshold value of a property (e.g., temperature) that was determined from the property's peak lateral gradient (front) location. The front is locally defined by an on‐board linear estimation, based on a single AUV's previous front crossing locations within a synoptic range.…”

“…Researchers have made substantial progress in enabling one AUV to autonomously identify and track various oceanographic features, such as upwelling fronts, the depth of the thermocline, and phytoplankton patches (Cruz & Matos, ; Godin, Zhang, Ryan, Hoover, & Bellingham, ; Petillo, ; Petillo et al, ; Petillo, Balasuriya, & Schmidt, ; Zhang et al, ; Zhang, Godin, et al, ). Collaboration between multiple feature‐detecting AUVs would greatly enhance synopticity, coverage, and endurance of automated oceanographic studies (Petillo, ; Petillo et al, ; Reed & Hover, ). The complexity and low bandwidth of the underwater acoustic communications channel poses challenges to multi‐AUV collaboration; however, those challenges can be overcome (Kemna, Caron, & Sukhatme, ; Kemna, Rogers, Nieto‐Granda, Young, & Sukhatme, ; Schmidt, Benjamin, Petillo, & Lum, ).…”

Autonomous & adaptive oceanographic front tracking on board autonomous underwater vehicles

“…1. Further details regarding the implementation of this multi-AUV front tracking behavior are found in Chapter 5 of [19].…”

“…The AUV's desired position on the helix at the current time, (x, y, z, tnow), is calculated considering the constraints above. Details of this calculation can be found in Chapter 5 of [19].…”

“…In this paper, a couple of novel methods for environmentally adaptive autonomous sampling and tracking along an ocean front are proposed and implemented using AUVs by employing the Autonomous Adaptive Environmental Assessment (AAEA) and Feature Tracking method developed by Petillo et al [18], [19]. AAEA is described as "a process by which an AUV autonomously assesses the hydrographic environment it is swimming through in real time.…”