Underwater (Marine) Robots Panel Notes

On 19th July Robotics Australia Group hosted a webinar on Underwater robots featuring an expert panel including: Adam Chadok, CEO and Founder of MetaSpatial Solutions, Ruth Patterson, Founder of Entira Oceans, Ken Chaddha, CEO and Founder of Yellowfin Robotics and Prof Stefan Williams, Professor of Marine Robotics and University of Sydney.?While we often refer to AUVs (Autonomous Underwater Vehicles) the webinar covered all forms of robots in the marine domain, including robots that operate on the surface of the water.

The panel spoke about a range of applications that marine robots were currently being used for in the Environment, Defence, Transportation, Renewable Energy and Emergency Services sectors. With Australia and our Pacific neighbours being responsible for 20% of the world’s oceans, the use of robots for ocean observation and exploring this still relatively unknown domain was emphasised, with the big opportunities of the future seen as a development of improved sensors to expand their range as well as platforms and algorithms to collect data and turn it into actionable insights about our ocean and the weather.

Some of the applications of robotics mentioned by the panel included:

● Mapping and physical dredging of shipping channels to certain depths

● Assisting with ocean wind farm construction

● Micro robots to support the navy, e.g., the ghost shark an autonomous robotic undersea warfare vehicle being designed and manufactured in Australia for the Royal Australian Navy

● Exploration for offshore resources

● Surface marine robots to collect data to fill gaps in our geographic knowledge and the opportunity to develop digital twins of our ocean.

Some of the technical challenges to be overcome in the marine environment include the need for robustness and persistence, designing electronics and their housings to be resistant to saltwater and to be in use for long periods of time without human intervention.?The platforms need to be able to operate and navigate in GPS-denied environments. Sensors continue to be developed, as well as change detection algorithms, with optical sensors not very good past 45m depth (blue spectrum) and acoustic systems deployed when in range of the sea floor to aid navigation and allow subsea obstacles to be avoided. Between 45m depth and the sea floor - not a lot of sensing is available, hence a lot of gaps in our knowledge. Continuing focus on optical sensors to be able to obtain high-quality imagery. Another challenge is cost - the small volumes of sensors currently deployed in the marine environment make them comparatively expensive.?This is a challenge to overcome for the provision of cheap reliable persistent marine robots.?Downstream advances in machine learning and artificial intelligence are allowing better processing of data, which increasingly needs to happen on the edge (on the marine robot itself) to allow both precise navigation and to detect any gaps in data as missions are underway, rather than these gaps being discovered after the mission during post-processing, and hence requiring the robots to be redeployed.

A major barrier is underwater communications. Getting platforms connected is a huge challenge. Even accessing standard communications between each Australian marina is an issue and hence most subsurface robots are tethered (for air and power and retrieval), which limits their effectiveness.?There is still a long way to go to make Remotely Operated Vehicles (ROVs) untethered. On the plus side, there is a better recovery rate of vehicles than had been possible in the past and there is some exploration of the use of “dispensable” robots, that can be deployed over long periods with no intention of retrieval. It was suggested that advances in space would be facilitated by better connections between robotic platform manufacturers and specialists in underwater communications and peripheral services, rather than expecting all the expertise to reside in a single company.?This would allow the expansion of the capabilities of many robot platforms.

As Australia’s robotics industry is very small there is a strong need to collaborate. Unfortunately, there is no support for local start-ups and so many head overseas. There is also a capability gap in the younger generation as there are few oceanography focussed qualifications, the main qualifications available are in marine biology and there are few robotics graduates who have domain expertise in oceanography. The lack of cross-disciplinary expertise is compounded by the competition for graduates with mechatronics degrees, who are in demand across all industries. The salaries for graduates with machine learning and artificial intelligence expertise are difficult for start-ups to compete with and hence access the talent they require. The lack of investment in start-ups based in Australia contrasts with the generous incentives and market access opportunities of countries like Abu Dhabi. The current focus on start-ups as potential unicorns (companies with a market capitalisation >$100m) detracts from the opportunities of building sustainable businesses that are not looking for a quick exit.

A notable success in marine robotics has been the Argo float program funded by the US National Science Foundation. The program involved the deployment of 4000 low-cost (<$15k) robotic "floats" equipped with a range of sensors deployed across the world's oceans, collecting data from the sub-surface and transmitting the data to Jason Earth-observing satellites. The platforms are designed not to be recovered and have revolutionised oceanography by giving persistent data on long-range ocean circulation and being much cheaper than trying to collect the same data using ships.

Finally, the panel agreed that as well as support for robotics start-ups, there needed to be support for all the ancillary industries supplying the robotics value chain, otherwise most components are sourced from overseas. The general sense from the panel was that robotics was a largely unrealised opportunity in Australia, that could be exploited with a structured approach to support from our government.

Watch the webinar HERE.