Requirements for Autonomous Underwater Vehicles (AUVs) for scientific data collection in the Laurentian Great Lakes: A questionnaire survey

Abstract

Using mobile environmental monitoring can aid in gathering ecological data to meet fish community goals in the Great Lakes. One such approach is the use of large Autonomous Underwater Vehicles (AUVs) to gather data, or the potential use of AUV swarms, where multiple small AUVs work together with each having different data-gathering capabilities. To understand data needs that could be collected by mobile sensor networks to inform decision making, we surveyed Great Lakes professionals involved directly and indirectly in such decision making. Basic data that respondents chose as most important to collect were water temperature, dissolved oxygen, chlorophyll a, turbidity, and blue-green “algae”, which seems to align with variables affecting fish directly or indirectly (through identification of harmful algal blooms). Specialized data chosen as most important were mapping of habitat characteristics, sonar of groupings of fish, and images/video. The time of year to collect all data was chosen as all seasons by the majority of respondents, the frequency most chosen was once a season for mapping of habitat characteristics, once a week for sonar detection of groupings of fish, and once per day for images/video and water temperature. Results were very similar when respondents were asked where data should be collected in the Great Lakes (i.e., tributaries, nearshore areas, etc.) except respondents indicated that images/video should be collected most in fish spawning habitats. Understanding data important to inform decisions of resource professionals will help guide the design of mobile and stationary sensor networks in the Great Lakes.

Introduction

As ecological research and practice embark on a new era steeped in expansive multidimensional data, new approaches and technology are required to effectively capture and leverage its potential. A basin-wide fishery research priority of the Great Lakes Fishery Commission (GLFC) is to identify what attributes of aquatic habitats are essential to achieve environmental and fish community goals and objectives and what methods should fishery managers use to categorize, prioritize, and inventory specific aquatic habitats (GLFC, 2017). We suggest here that decision makers could apply robotics to collect attributes of aquatic habitats and inventory aquatic habitats essential to achieve environmental and fish community goals and objectives. Robotic sensor networks find increasing use in environmental monitoring as they can collect data from obscure, hard‐to‐reach places over long periods of time (Tokekar et al., 2010). Collection of data at unprecedented spatial and temporal scales provides capability crucial in understanding complex phenomena that, in turn, are invaluable for scientists tackling critical environmental issues (Tokekar et al., 2010).

An Autonomous Underwater Vehicle (AUV) is generally described as being capable of operation in the absence of a human controller; thus, it must be able to make real time decisions to traverse the aquatic environment on a desired path through autonomous navigation and control (Kumagai et al., 2002). In research, AUVs are commonly used to carry hydrological, geophysical, and/or biological sensor payloads (Mora et al., 2013). This automation of observation, data collection, and data preprocessing represents a paradigm shift in ecological science because data collection that was deemed infeasible or impractical due to inherent dangers, economics, and sheer tediousness are now on the horizon due to advances in enabling technologies. The design of AUVs is improving rapidly. For example, Tantan is an autonomous underwater vehicle (AUV) developed to monitor the quality of water in lakes (Kumagai et al., 2002) by carrying multiple sensors to monitor the distribution of plankton and to measure the level of dissolved oxygen in the water. An AUV equipped with a high‐resolution stereo‐imaging system (Williams et al., 2009) was used to obtain images to study the nocturnal camouflage behavior in cuttlefish through various short missions over six nights.

AUV-supported research in the Great Lakes is expanding and has included the use of a long-range AUV deployed in Lake Michigan (Bennion et al. 2019), studies of Lake Michigan river-mouth mixing zones (Jackson and Reneau, 2014), a buoyancy-driven ocean glider in Lake Superior (Austin, 2012, Austin, 2013), and AUVs for lake-wide monitoring in Lake Ontario (White and Boyer, 2013). The use of an AUV or AUVs can provide significant advantages over traditional sampling techniques. For example, a survey of outer Milwaukee Harbor using an AUV required less than 7 h for approximately 600 profiles compared to the 150 h it would have taken using traditional methods in a manned boat (Jackson and Reneau, 2014). One current application in the Great Lakes, the Real-time Aquatic Ecosystem Observation Network (RAEON, https://raeon.org/), is able to collect an extensive amount of data on water quality and other parameters using multiple stationary buoys and two autonomous sub-surface gliders.

AUV swarms can complement the existing sensor network in the Great Lakes by providing more mobile data collection capabilities. AUVs in swarms are easier to deploy as they are designed to be smaller, more maneuverable for use in coastal and riverine environments, and lower in cost and less disruptive to wildlife than larger AUVs. The idea of using multiple cooperating AUVs (a swarm) was first considered in the 1980s (Blidberg, 2001), garnering inspiration from social animals (Kube and Zhang, 1993). As opposed to solitary behavior, swarm occurrences in nature exhibit the superior sensing capabilities associated with large groups and achieve improved results through distributed workloads. AUV swarms can also be scaled to the magnitude of the task. Kube and Zhang (1993) indicate that optimal swarms are: 1) scalable to the desired mission, 2) resilient and thus able to adapt to loss or malfunction of some elements of the swarm and be tolerant to external threats, and 3) flexible and thus should respond well to changes in the environment or mission requirements. Allison et al. (2019) tested individual AUVs for a swarm which required no special equipment to deploy, were only 10 kg with dimensions of 46 cm × 34 cm × 26 cm, and cost approximately $3000 (2017$). Larger AUVs can also be used in a swarm to carry very heavy sensor payloads as needed. In a preliminary scaling experiment, they employed a multi-agent simulation platform assuming the equivalent of 0.125 km³ for the search space for each AUV in the swarm (Fig. 1). They found that the swarm functioned optimally with a swarm size between 10 and 340 in ideal conditions, and 100 under field conditions, with a maximum of 12 AUVs assigned per base. Each base can move with the AUVs and is a solar-powered charging station that collects data from the deployed AUVs that can be sent wirelessly to a user. We are optimistic with the performance and scaling consideration of the swarms as the supporting technologies are becoming cheaper and more efficient while sensing is becoming lighter and more accurate.

The collection of ecological data using multiple AUVs promises the potential to collect multiple types of data in nearly all aquatic environments to address second–generation ecological questions (Rundel et al., 2009). Tokekar et al. (2010) built a robotic raft for monitoring tagged carp in Minnesota lakes as a proof-of-concept design to successfully localize tagged fish through the collection of signal strength and bearing measurements. A small number of lightweight robotic rafts would be ideal as they can autonomously reconfigure themselves based on the location of tagged fish (Tokekar et al., 2010). A swarm of multiple, inexpensive AUVs working together can be designed to collect data in the Great Lakes, with each AUV carrying different sensor payloads to collectively achieve relatively comprehensive data coverage.

The overarching question to guide the design of AUVs to address environmental and fish community objectives is what type of data should be collected, when, how often, and in which types of habitats in the Great Lakes ecosystem? While the autonomic software framework for AUV swarms, for example, has already been designed and tested (Allison et al., 2018, Allison et al., 2019), the questions regarding what type of hardware and payload is needed to collect the relevant data had remained unanswered. The motivation for integrating decision makers and other primary stakeholders in the design process is grounded in requirements engineering and user needs analysis (Lindgaard et al., 2005). Determining the needed hardware or sensor payloads to acquire the most relevant data on fish communities in the Great Lakes is best left to the fishery professionals and stakeholders of the fish communities of the Great Lakes. Technological innovations concerning the application of AUVs to the ecosystem is vast and advancing in a very rapid manner. Therefore, we surveyed some fishery professionals and stakeholders to identify their data needs for the Great Lakes that could potentially be met using multiple AUVs working together (a swarm). This data will be useful to other researchers designing or using stationary or mobile sensor networks in the Laurentian Great Lakes to ensure data may be most useful to inform decision-making.