How Technology Enables Innovation in a Towed Vehicle System

The following was adapted from an article by Global Oceans’ CEO Jim Costopulos, published in ROV Planet Magazine, Issue 26, 2021

The original Ocean Explorer 6000, formerly owned and operated by Oceaneering International, is a 2-body, passive towed system with a depressor and side scan sonar. Global Oceans will be converting this vehicle to an actively-controlled, maneuverable, single-body system, called the Innerspace 6000 TIA (Towed Instrument Array) that will be powered from the surface and will dynamically communicate with the winch system to control umbilical management.

The new Innerspace 6000 TIA will incorporate innovative advanced technology to enable precision attitude control and vehicle steadiness (high-precision INS with acoustic positioning and GPS), precision altitude control (+/- 0.3 meter), geodetic positioning (to < 10-meter CEP), Multibeam Echo Sounder (MBES), embedded magnetometer, autonomous behavior and fault management, and system redundancy. Vehicle exostructure will also be modified to enable actuated control fins (Figure 1).

The redesigned system will solve several problems inherent with towed vehicles of this type. For one, a passive Towbody acquiring data at depth requires a long vessel turnaround time when conducting gridded transects, between five to eight hours, during which the vehicle is usually unsteady and unable to maintain speed during the turn, leading to an inability to collect data during this phase of the transect.

Maintaining Vehicle Steadiness & Power Supply

Maintaining steadiness and heading with these systems is difficult. In the case of a towed vehicle operating at a depth of 6,000 meters, with cable deployment of three to five times depth, the vehicle will be at least 18,000 meters or more away from the vessel and keeping the vehicle safe and off the bottom during a turn is a challenge. With a control system, altitude can be maintained during the turn and the steadiness of the system required by sensors that collect data can be maintained. This means the vehicle can collect data through the entire turn, through a smaller radius, in a shorter period of time.

As an actively-controlled system, the new vehicle will essentially drive itself, sending signals back to the winch from the vehicle’s control system to dynamically control umbilical deployment, so the vehicle controls the winch from its vehicle control system. This configuration enables performance superior to a passive Towbody and has advantages over an AUV.

Sending power down to the Innerspace 6000 TIA eliminates the need to charge batteries on long surface intervals. The new vehicle will not need to surface to maintain full power and can run at depth for an extended period collecting data. The new system will deliver its data stream in real time, for analysis in real time, compared with an untethered AUV which must surface and come aboard to offload data, review the data, and to program the next sortie. A surface-powered vehicle also means instrument power consumption is not an issue compared with battery-powered systems. We can run more instruments and sensors on full continuous power without regard to onboard power consumption.

Vehicle-Driven Navigation and Control

The TIA system will also have navigation capability. A redesigned exostructure with actuated fins, buoyancy control, and an internal control system, coupled with hydro-dynamic control, will eliminate the need for a depressor body. The new vehicle will be converted to a single body system for more simple launch and recovery with a standard A-Frame crane or other handling system. For shallower deployments we can add the depressor for improved decoupling from ship motion if needed. By linking the forward-looking sonar with active control of the vehicle, the risk of losing the vehicle at depth is low relative to other technologies.

Extensive computational modeling for redesign of the TIA has been conducted with a physics-based model of vehicle hydrodynamics and cable dynamics in six-degrees of freedom, incorporating wave effects, variable water density, variable currents, and other factors. Each individual component and instrument placed on the vehicle can be modeled separately to predict its hydrodynamic impact.

Superior Productivity in a Towed System

A comparative productivity analysis compiled by our design team on relative performance between a new actively-controlled Innerspace 6000 TIA, a passive towed sensor system, and an AUV provides a quantitative estimate of Operational Time (project-deployed time when the vehicle is generating environmental data) vs. Non-Operational Time, or “operational overhead” (project-deployed time when the vehicle is not generating data).

For this comparison, we designated bottom time, coverage area, deployment, and maintenance time (including ascent/decent, topside data processing, maintenance and battery recharging turnaround, OPS planning, Launch & Recovery, and estimated turn time at depth during which the vehicle is not gathering useful data).

The calculation assumes AUV bottom time increments of 24 hours before ascent for maintenance, data download, etc. and redeployment (not including possible AUV ascents/descents for taking GSP bearings); and towed system bottom times of 240 hours (for both passive and active towed systems) before coming to the surface; then running the model for 30 sorties. Calculating average total percent of operating vs. non-operating time for each type of system resulted in the productivity estimates shown in Figure 2.

Roughly half of the operational time for the AUV is devoted to non-data-gathering activities. The passive Towfish improves on this performance by about 20% but remains burdened by the inefficiencies of passive-body management mentioned above. The significant productivity gains shown for the actively controlled towed system, reflected in the technology strategy for the new TIA vehicle, stem primarily from two factors: the ability to conduct continuous data collection through the turn on a gridded transect, and the continuous full-power supply available from the surface, without the need for battery recharging, data downloading, and re-programming on long surface intervals. This enhanced efficiency translates directly to cost savings and greater capacity for a range of deployed sensors and instruments.

A particular cost advantage of the new TIA system will derive from coupling the capacity for high-resolution bathymetric mapping at depth utilizing the on board MBES, with deployment on MARVs as project vessels. Given that MARVs are assembled as science-mobilized time-chartered OSVs, which have become largely commoditized from a cost standpoint, this approach will enable lower-cost seafloor mapping as a function of data output, resolution, and quality compared with using high-cost survey vessels (See more about Global Oceans' MARV model here).

AI and Machine Learning Integration

The ability to supply continuous, direct power to the vehicle from the surface will also enable long-range deployment of high-powered lighting and video or photographic systems for deep-sea benthic imaging. This capacity will benefit two projects launching from Global Oceans, the Innerspace Deep Sea Initiative to conduct benthic imaging surveys and biogeochemical sampling through extreme environments such as hydrothermal vents and hypersaline pools; and the Global Seamounts Project which will seek to generate new biophysical behavioral models of seamount ecosystems. 

For example, for the seamounts project we will conduct visual transects with the Innerspace 6000 TIA across abyssal regions adjacent to seamounts to generate geomorphic proxies for biological activity, including for biomass and biodiversity. To achieve this, a statistical machine learning process for pattern recognition will be employed to analyze wide-area visual data, coupled with analysis of representative core samples to generate a set of training data for the AI.

A deep-neural network, or perceptron, will be developed to relate biophysical proxies to biological indicators. Once developed, classification algorithms will enable real-time in situ predictions of biological characterization of the benthic seabed utilizing an edge-computing device to be installed on the survey vehicle.

The new Innerspace 6000 TIA system will complement a range of tethered and autonomous vehicles and benthic lander systems to be deployed on numerous expeditions and future projects.