Fast paced developments in the field of Marine Robotics have led to tremendous progress in a vast array of topics that include sensors and energy devices, communication techniques, vehicle hydrodynamic design and propulsion, and advanced systems for mission command, navigation, and control. This flurry of activity is consequence of the formidable challenges set for in the context of scientific and commercial mission scenarios that include, among other, fast seabed surveying, marine habitat mapping, inspection and monitoring of critical underwater infrastructures, and the execution of demanding tasks at various ocean depths.

Notwithstanding the remarkable results achieved, both from a conceptual and technological standpoint, there is an urgent need to develop a new breed of hybrid marine robots capable of transitioning between multiple modes of operations in a seamless manner, in reaction to on-line detected events, ranging from tele-operation to co-control and fully autonomous modes, as a means to intelligently include a human operator in the loop, yielding adaptive autonomy capabilities. A compelling example is the one where the vehicle starts by executing a fast survey of the seabed in a fully autonomous mode, after which, based on the analysis of the data acquired, a human operator may trigger the execution of a new mission in a tele-operated mode for close inspection of conspicuous features, followed by the activation of a co-control mode. The latter allows the operator to focus on a specific task (e.g. sampling the seabed) while the robot executes the complementary task of holding a desired position with respect to a conspicuous point detected by an acoustic or vision imaging system.

In this context, in line with current trend worldwide, we envision the development of robotic-based systems capable of operating in fully autonomous mode, while allowing for human intervention in tele-operation or co-control modes based on real time information provided by AI based sensing and environmental classification tools. Keeping in mind these challenges and the required technological advancements, this joint project proposes a number of innovative developments for marine robots and related systems with a view to safe, reliable, and effective execution of scientific missions by the partner group in the area of marine ecosystem studies using proprietary vehicles of NIO and IST-ID and advanced AI based systems for habitat characterization of the Univ. Aveiro. Specifically, we envision the design, implementation, and testing of:

– Advanced hybrid underwater vehicle guidance and control primitives and subsystems that address explicitly the requirement for seamless transition between autonomous, tele-operated, and co-control modes of operation of marine robots through the use of effective thruster allocation methodologies and sensor-based control strategies

– Artificial Intelligence (AI) tools to enhance object detection and classification capabilities and to extract relevant information for marine habitat characterization or the assessment

of living and non-living resources. Numerous applications require underwater seabed imaging, which is normally achieved through acoustic sensors or at times through vision based systems. A blend of the two methods using advanced sensor fusion techniques can provide richer information and an improved, holistic image of the underwater realm of interest.

– Robust software-defined mixed aerial, acoustic, and optical communication networks as a means to guarantee reliable robot-base station communications in autonomous mode, while enabling fast data downloading from an underwater station to the robotic vehicle in al modes of operation.

The project will witness full integration of selected subsystems on robotic vehicles that are property of the groups in India and Portugal, followed by the execution of representative field tests. The proposed work plan leverages on previous strong collaborative links established between CSIR-NIO and ISR-IST initiated in 1999. For more than two decades, the two

institutions have exchanged researchers periodically, participated in projects and research actions that led to the development of the MAYA, Autonomous Vertical Profiler, and

MEDUSA AUVs, and carried out experiments with marine robotic vehicles in Goa and the Azores. A representative landmark was the development of the MAYA AUV, that received partial support from DST and the FCT under the Indo-Portuguese Research Cooperation in S&T ( 2003-2006). The Portuguese PI Antonio Pascoal is an Adjunct Scientist at NIO, and

in this quality he has teamed up with NIO scientists to organize with tremendous success a series of Marine Robotic Schools in Goa over the past years. Presently, NIO and IST are partners in the EU funded ECOBOTICS project. The Univ.Aveiro has a long standing tradition of cooperation with ISR-IST and brings into this project the much needed expertise