Optical Sensors: Enabling Explorers to See the Depths Like We See the Surface

Optical Sensors: Enabling Explorers to See the Depths Like We See the Surface
Wide-area 3D seabed data captured from the Voyis Insight Pro laser scanner. (Image credit: Voyis)

Human ingenuity has led to a world of discovery, yet our ocean’s deepest secrets remain shrouded in darkness. The deep sea covers approximately 65% of the Earth’s surface and a has remained largely undocumented, with surveys being limited to low-resolution mapping from surface vessels and satellites, and localized observation with submersibles.

Efficient wide area optical exploration in deep water is key to uncovering the ocean’s mysteries and is vital for documenting and managing the habitats, archaeology, and critical infrastructure that call the ocean home.

MAKING THE DEEP SEA MORE ACCESSIBLE

Today we rely on robotic underwater vehicles, equipped with powerful sensors, to get us to these remote regions in order to survey large areas of the deep sea. These vehicles are constantly evolving, gaining new capabilities that enable us to travel further than before.

Whilst crewed submersibles have allowed humans to visit depths of over 4000 m, they are complex, high risk, and costly to operate, making them ineffective for expansive surveys. Remotely Operated Vehicles (ROVs) eliminate the human presence to reduce costs and improve safety, but their slow speed also makes them inefficient for capturing wide-area datasets. The ideal platforms are Autonomous Underwater Vehicles (AUVs) which can dive to depths of over 6000 m, travel at speeds up to 6 knots, and can carry a wide range of sensor payloads.

The concept of using AUVs for wide area exploration was demonstrated by Ocean Infinity with their HUGIN AUVs in their search for the MH370 aircraft wreckage. During their search they noted that their fleet emitted 72% less CO2 on average and covered 125,000 square kilometers of seabed in just 138 days—a far cry from the 837 days it would take typical survey vessels. With this incredible achievement it is evident that AUVs are enabling researchers to access the deep, at a lower cost, and with a reduced environmental impact. But getting there is just the beginning. Having tightly integrated, high-resolution sensors onboard is paramount to improving our understanding of the world’s last frontier.

GAINING A DEEPER UNDERSTANDING

One of the main challenges with documenting and observing the deep sea is the trade-off that is typically made between coverage and resolution in subsea instruments. While some sensors can capture data across large areas, they do not have the resolution required for gathering quantifiable and actionable data.

For example, AUV side-scan sonar can achieve incredible coverage rates, mapping over 200 m on each side of the vehicle, but their limited resolution is best suited to detect objects, not to identify and interpret them. Conversely, optical sensors like 3D laser scanners and stills cameras can provide highly detailed quantitative and qualitative datasets but are not well suited to mapping the entire ocean when the typical image only has a 10 m swath. To achieve both coverage and resolution the sensors and vehicles must work together as a tightly integrated system, with acoustic sensors first detecting features of interest and then optical sensors reacquiring the target for assessment with higher resolution data.

At Voyis, we are looking to overcome these challenges with our Recon AUV payloads that provide both 3D laser and high-resolution color image datasets simultaneously. This pre-configured module is outfitted with Voyis’ Insight long-range laser scanner and Observer and Nova Imaging system which includes our new 6000 m rated color stills camera. Wavefront’s side-scan sonar can be integrated into the Recon payload for a complete package that achieves both wide coverage and high resolution.

ENABLING SENSOR AUTONOMY

With these AUV payloads it has never been easier to remotely collect incredible data on both small and large platforms. However, another core challenge is finding an efficient way to conduct this exploration and review the vast amount of data.

An AUV operating in the deep ocean cannot transmit large datasets back to the surface, and therefore must be evaluated by an operator once the vehicle is recovered. For wide-area surveys this process can take weeks and usually means that the surface vessel has moved on to different location. When looking to combine sonar and optical data, the window of opportunity for an optical survey of a detected feature has usually passed. It is therefore paramount that we look to onboard data processing for the automatic detection of targets of interest to drive in-mission optical reacquisition, for a more efficient workflow.

With the integration of EIVA’s Onboard workflow Manager and Deep Learning into Voyis’ Recon payloads, this type of unsupervised decision making is now achievable. Both side-scan and optical data can be analyzed onboard the payload in real-time, identifying anomalies, perceiving the environment, and providing information to the vehicle that can enable true autonomy.

SEEING THE DEPTHS LIKE WE SEE THE SURFACE

When taking photos underwater at long ranges, water absorbs the visible light spectrum unevenly causing color images to appear mostly blue or green when in reality the ocean is filled with color. Voyis is making it possible to see the depths like we see the surface by instantly enhancing and correcting images as they are collected. Our robust machine-learning-based color correction algorithms help us to see consistent, true-color photos, that enable effective analysis of marine ecosystems.

When combined with navigational data, it is possible to deliver consistent images across incredibly long surveys. This uniformity produces wide area consistent image mosaics without variations in quality from image to image. Broad regions can be observed all at once to gather a better macro understanding and then zoomed in on to clearly see tiny details of your subject. Adding onboard analysis software, like EIVA’s Deep Learning, enables the automated detection of marine life and seabed features, further accelerating the time-to-results.

All the recent innovations in platform and sensor technologies have made it possible to explore vast areas of our deep oceans more efficiency and with better data. The final step is putting it into action on a wider scale to enable further discoveries, which will require researchers, surveyors, vehicle designers, and sensor manufacturers all working together to Illuminate the Unknown.

To learn more visit www.voyis.com or contact sales@voyis.com.

This story was originally featured in ON&T August 2021. Click here to read more.

Image

Corporate Headquarters

Ocean News & Technology
is a publication of TSC Strategic

8502 SW Kansas Ave
Stuart, FL 34997
info@tscpublishing.com