The OCEANS conference was slated to be held in Singapore and subsequently in Gulf Coast, USA for the year 2020. However, due to the COVID-19 pandemic, IEEE Oceanic Engineering Society and Marine Technology Society decided to combine the two in-person conferences into a single virtual conference called as "Global OCEANS 2020: Singapore – U.S. Gulf Coast". The virtual conference was online (on-demand) from October 5 - 30, 2020 with the live virtual part taking place from October 5-14, 2020.

One of the technology panel discussions held as part of this exciting event was the "Autonomous Systems: Emerging and Future Technologies Panel". The main goal of the discussion was to identify challenges and opportunities in emerging technologies such as ocean sampling networks, underwater IOT, resident underwater vehicles, underwater human robot interaction, and ultra-long endurance systems. We had an experts’ panel comprising of

1. Hanumanth Singh, Professor at the Northeastern University, USA,
2. Tom Curtin, Senior Scientist at the Applied Physics Laboratory, University of Washington, USA,
3. Nikola Mišković, Professor at the University of Zagreb, Croatia and,
4. Ralf Bachmayer, Professor at MARUM, University of Bremen, Germany.

The panel was moderated by Mr. Richard Mills (from Kongsberg), and chaired by Dr. Bharath Kalyan from National University of Singapore.

To start off, Prof. Singh briefly touched upon four cutting edge topics:

Networked systems: An illustration of different modes of communication amongst underwater, surface, and aerial vehicles.

• Autonomous oceanographic sampling network (AOSN) involving networked systems comprising of underwater vehicles, docking & charging stations, underwater & satellite communications, and, adaptive mission planning;
• Haptic fine scale manipulation using lighter and more flexible manipulators mounted on Autonomous Underwater Vehicles (AUVs) and Remotely Operated Vehicles (ROVs).
• Fisheries independent stock assessments, where robots/fixed cameras can count fishes instead of catching fishes by trawlers for stock assessment. The technologies developed in this context could also be used for monitoring habitat around windmills and aquaculture farms.
• Underwater Internet of Things (IoT): Like factory-based automation, can we build an underwater infrastructure encompassing moorings, underwater vehicles, gliders, ships with all talking to each other over a single communication framework? Prof. Singh summed up saying that with all the latest technological developments, the time is ripe for above ideas to come to fruition.

Moving towards ubiquitous AUVs

The need for a network of infrastructure was also highlighted by Dr. Curtin. He pondered over the technological advances needed for the current generation of AUVs to become truly ubiquitous (resident) in the world's oceans including in-situ infrastructure that could allow AUV recharging and improve AUV navigation. Dr. Curtin suggested to look into and learn from the land and air transportation community where large support infrastructure has facilitated the smooth transition of these transport systems. For instance, the full Federal Aviation Administration (FAA) VHF Omni-Directional Range (VOR) navigational network infrastructure has made today's commercial aviation possible. He also mulled upon leveraging on technological developments that weren't purpose-built for underwater use like Iridium and GPS, which have now been used heavily for AUV operations, when on surface. The need for truly open ocean infrastructure was identified by Dr. Curtin as key to make ubiquitous AUVs a reality. Some examples can be found among the existing infrastructure such as ARGO floats, global tropical moored buoy array, international arctic buoy array and existing undersea cable network. He highlighted the need for a robust academia-industry collaboration to achieve this vision of open ocean infrastructure.

Human-Robot interaction

Prof. Mišković talked about the challenges facing the underwater technology community over the next decade. He compared the developments underwater against "Industry 4.0", which was primarily aimed at improving efficiency in factory settings using interconnected devices.

With the use of multiple cooperative underwater vehicles working through Underwater IoT communications and achieving pre-defined tasks, Prof. Mišković believes that "Industry 4.0" is already achieved underwater and we are now in the era of "Industry 5.0" where humans and robots would be capable of working in tandem.

Several projects are working on Human-Robot interaction underwater including a pan-European project which is looking into diver interacting with underwater robots and performing tasks on instruction.

More challenges and opportunities

Launch and recovery of AUVs and ASVs in different sea states can be challenging.

Prof. Bachmayer talked about scenario-based view on autonomous marine platform developments with emphasis on challenges and opportunities. The first challenge pointed out by Prof. Bachmayer was the launch and recovery of AUVs and Autonomous Surface Vehicles (ASVs) in varying sea states. This is the most tricky part of vehicle operations and thus there is a need for technological improvements. He provided an example of such an improved operational scenario involving an ASV launching and recovering an AUV. Prof. Bachmayer also highlighted the challenges on long-endurance surface and underwater marine platforms, sensor stability and calibration, marine antifouling strategies without the use of bio-toxins among others. The high cost and complexity of navigation sensors and other sensing payload solutions are as well detrimental to the development of AUVs. Another need identified by Prof. Bachmayer was the need for more intelligent and adaptive fault-response to increase systems’ robustness.

Nonetheless, like Prof. Singh, he also believes that with technological advancements, there is great potential in the following areas:

• Soft robotics: the use of soft sensors/actuators can reduce drag, provide seamless actuation and has the potential to increase the overall vehicle robustness although it can increase control complexity and decrease efficiency;
• Multi-modal sensing & communication – One of the interesting research aspects for the future is how to take advantage of multi-modal sensing and communication, i.e., how to use acoustic and optical communications modes in a collaborative fashion and switch between the modes when there is an environment induced bandwidth reduction?
• Environmental sensing – One of the future advances in environmental sensing is related to the sampling coverage. For instance, for water property measurements along the water column, can we do area or volumetric measurements instead of the widely practiced line measurements along vehicle's trajectory?

Finally Dr. Bachmayer, like all the other panellists, pondered upon the variety of available systems such as moored buoys, profiling floats, underwater gliders, wave gliders, ships, AUVs, ROVs and the need to network all these systems to bring the data together, merge and assimilate the data into numerical ocean models and environmental models. He believes that there should be higher effort towards producing relevant and meaningful data as they would be always sparse when compared to the size of our oceans. Connected to this effort, there is also a need to enhance talent development and capability building and a need for working with local communities.

Q&A session

After a short presentation by the panellists, a live (and lively) Q&A followed. Among the questions raised by the attendees (and moderator) was - "What can the community can do to enable higher usage of underwater robots?". In other words,

...can we lower our entry barrier and make underwater robots more accessible? One possible way highlighted by Prof. Bachmayer was the use of telematics. It is currently in use in the Oil & Gas industry, primarily for diagnosis and could potentially be used in other areas. Remote presence would enable scientists to stay onshore thus significantly reducing the offshore operational workforce.

Another point raised by Prof. Singh was the complexity of AUV operations in comparison with other vehicular technologies such as Unmanned Aerial Vehicles (popularly known as drones), that are being adopted much faster due to their ease of use. Prof. Singh also recalled the lack of infrastructure (as mentioned by Dr. Curtin) and the lack of standardization in the software used by underwater robots. While in the past, every research group had their own code, the current status is much better, helped by the popularity of middleware such as the Robot Operating System (ROS). Lack of standardization was also mentioned by Dr. Curtin, who further suggested to look into other fields such as aviation to check on how the standards were developed and agreed.

Iridium satellite communications is often used in autonomous systems at sea to send/receive compressed data.

Another question from the audience pondered on the limitations of Iridium satellite communications with respect to bandwidth and cost. According to Prof. Bachmayer, while the cost is getting lower, the question is not about either to use Iridium or other satellites. The main point is that prior to Iridium there was no good alternative. While the bandwidth is in the order of tens of kbps, it is enough to send compressed data (including video).

Swarm robotics was briefly addressed by Prof. Miskovic. He identified the biggest challenge to be the logistics of unmanned operations. He opined that under high sea states, launch and recovery of a single vehicle is already a challenging problem, and the operational complexity increases significantly with multiple vehicles and/or swarm.

As an example, he cited the EU project subCUltron, which aimed at the biggest swarm in marine robotics with dozens of robots belong to 3 types (one surface and two underwater). In this project the biggest challenge was the launch, recovery and monitoring (besides the large production) of the swarm systems, as the manpower requirement increases significantly with higher number of vehicles. According to him, we are still not paying enough attention to swarm systems in the marine robotics field.

The launch and recovery of multiple vehicles was also addressed by Prof. Bachmayer. The issue is that sometimes, by the time some vehicles are deployed, some are already at the surface ready to be recovered. Moreover, by increasing the time in the water, the weather window is decreased. This means that sometimes is not possible to recover the vehicles (while launching them is always possible) breaking what Prof. Singh calls the first law of field robotics: the number of recoveries should equal the number of deployments.

A question on the use of multi-domain multi-vehicles systems involving UAVs, ASVs and AUVs was answered by Prof. Miskovic. He mentioned that several challenges occur in such systems. On one side, software frameworks are different, not only among different domains, but sometimes within one domain. Further to this, there are obvious physical differences in the medium of communication and thus the bandwidth of operations, making multi-domain multi-vehicle co-operation complex. Finally, there is a lack of cross-fertilization between the different sectors as some groups do only marine robotics, others only aerial robotics, so it can be hard to speak between niches.

Academia-Industry engagement

Mr. Mills, the moderator had a question to the panellists underpinning the challenges on engagement betwen academia and industry.  Should academia do a better job to look in the market before developing their vehicle or should industry market better their products?

Both Prof. Singh and Dr. Curtin highlighted the differences in culture between academia and industry. One issue is that the AUV market is not big and AUVs are tools as Dr. Curtin answered so they need to be applied to a given project. The other issue mentioned by Prof. Singh is that while industry would love to have more academics working with them, Intellectual Property questions are raised sometimes. The advice he gives (and was given to him) was to not build yet another new vehicle (without a very good motivation). With the small market and the number of AUVs/ASVs already built, in many cases it doesn’t make sense to build another one that will be very similar to the ones existent in the market.

The panel concluded with a short question regarding what each speaker was most proud of, and what they were most excited about, looking forward. The panellists highlighted the significance and contributions of their students and other team members as their biggest achievement. They also expressed their urge to go back to physical meetings and conferences in the post COVID-19 pandemic world, as well as to use marine robots to address climate change.

Read more articles on OCEANS 2020 conference coverage or on UN Decade of Ocean Sciences-related coverage.

About the authors:

Bharath Kalyan is a senior research fellow at Acoustic Research Laboratory, Tropical Marine Science Institute, National University of Singapore, Singapore.

Fausto Ferreira is a leading researcher at Laboratory for Underwater Systems and Technologies, Faculty of Electrical Engineering, University of Zagreb, Croatia.

Image credits:

"Iridium satellite phone" by brunosan is licensed under CC BY 2.0

"130711-N-PO203-052" by Office of Naval Research is licensed under CC BY 2.0