With the expansion of global trade over the last few decades, shipping has undoubtedly become an integral part of the global economy. However, the shipping industry experiences more than 600 accidents annually of which more than 60% can be attributed to human actions. The recent incident of loss of heading of the ship Ever Given leading to the ship being wedged in the Suez canal is a clear example of such accidents. In addition, more than a billion dollars are spent annually to deal with pirate attacks against merchant vessels. These factors have served as a strong motivation to kick off a race to develop autonomous vessels. Currently, several countries including the United Kingdom, members of European Union, Singapore, South Korea and Japan are actively pursuing the development of fully autonomous ships and aim to develop a fully working prototype before 2030. However, India has been lagging in this global race and very little research has been performed in India in this direction.
The current research on Autonomous Surface Vehicles (ASVs) is largely focused towards small vehicles especially with a catamaran hull form. Most of the unmanned surface vehicle (USV) platforms commercially available are also catamaran shaped. These vehicles are particularly useful for applications involving surveying smaller areas and collection of water samples. However, these crafts exhibit different dynamics as compared to traditional ship hull forms. Consider one such example. The catamaran hulls are usually equipped with either propellers or water jet propulsion units on both hulls that allow the vehicle to turn using a differential thrust mechanism. The traditional vessels however have a single propeller and use a rudder to turn. The dynamics of rudder are significantly different from a differential thrust mechanism. These and other differences in dynamics of both hull forms result in some of the control approaches implemented on catamaran shaped hulls inapplicable to full scale ships. For the development towards full scale autonomous ships, research must focus on scaled models of real ship types.
The department of ocean engineering at IIT Madras has perfected the art of building scaled ship models for performing experiments to aid the design process of ships. However, very little research has investigated the development of free running models that can be controlled either remotely or autonomously by an onboard computer. Under this center, the first area of research would entail the development of scaled free running ship models that can be driven by an onboard system that can communicate with an onshore system or cooperatively with other models using WiFi or RF (Radio Frequency) based signals. The onboard system will be connected to several sensors such as GPS, inertial measurement unit, camera and lidars for navigation and sensing their environment. Development of such a platform is a non-trivial task due to the naval architecture, mechanical and electrical challenges associated with it and requires interdisciplinary collaboration. These models will act as platforms to further investigate various control algorithms.
The second line of research under the center will involve the development of newer and unconventional control strategies for autonomous surface and underwater vehicles. Apart from the fact that ships are under-actuated systems, the significant nonlinearities in the dynamics/hydrodynamics of the vessels pose additional difficulties in the implementation of the control algorithm.. The control algorithm needs to stabilize the system, ensure reference tracking, reject disturbances, attenuate noise and be robust to uncertainties and unmodeled dynamics of the vehicle. Such problems were in the past formulated as an optimization problem and resulted in the development of optimal control and model predictive control approaches. However, with the advent of artificial intelligence and machine learning, new approaches to derive a control law are being actively explored. Particularly, data driven methods will be explored for control of underactuated vehicles under this line of research.
The final broad topic involves the development of a software framework for autonomous mission planning and execution. Such frameworks are also known as robotic middlewares. Although there are several middlewares available today, they also have severe limitations with respect to resolving heterogenous data from different hardware systems, handling scarce communication in marine environments, holistic awareness of the environment while decision making etc. One of the popular frameworks today for robots is the Robot Operating System (ROS) that has gained popularity for both aerial and ground robots. However, for marine and underwater robots, there are still several challenges. Under this center, the development of a new framework based on open source architecture will be undertaken to plan and execute autonomous missions with particular emphasis on marine robots. The end goal of the center is to emerge as a collaborative hub where international and IIT Madras faculty can join forces to address the challenges associated with the development and deployment of autonomous vessels and play a key role in bringing autonomy to the oceans.
In house development of multiple free running ship models with an integrated on-board computer and sensors
- Use of these vehicles for research
- Use these vehicles for undergraduate teaching
- Use these vehicles for a graduate course on autonomy
Publications in high impact international journals and conferences on the following topics:
- Underactuated control of ships for path following, trajectory tracking and obstacle avoidance using traditional control approaches
- Underactuated control of ships for path following, trajectory tracking and obstacle avoidance using data driven control approaches
- Autonomous operation of a single vessel
- Cooperative autonomous operation of multiple vessels
Development of a software framework for handling control and autonomous missions of single and multiple vessels
Extensive collaboration with foreign faculty/experts for the development of the state of the art technology in the field of autonomous ships
Scaled ship models have been built and are being outfitted with onboard computers, propulsion and steering systems and sensors for unmanned operations. Some pictures and videos are provided.
- Dr. Srikanth Saripalli, Texas A&M University, USA Link
- Dr. Mike Benjamin, Massachusetts Institute of Technology, USA Link
- Dr. Prasad Pereira, University of Tromsø, Norway Link
- Dr. Antonio Pascoal, Instituto Superior Técnico, Portugal Link
- Dr. Sanjay Sharma, University of Plymouth, UK (Currently an MOU is being discussed) Link
- Seaconvoy Systems Engineering Private Limited - A collaboration with them is in progress to improve the build process of the hull and integration of the propulsion and steering systems.
External international site links for project
Center for Marine Autonomous Systems (CMAS) works towards the development of autonomous navigation of marine surface vehicles. Marine autonomy will result in a reduced operational cost of ships because of the reduced fuel consumption and crew cost. Lesser fuel consumption and reduced crew cost will result in a 60% reduction in the ship operating cost. Since 90% of the world trade happens through the seas, a reduction in operational cost will boost the economy by bringing down the transportation cost associated with goods.
Shipping accidents, particularly those associated with tankers, can result in a large-scale environmental pollution, extinction of marine ecosystems and a huge economic loss to the owner and marine transportation. It is estimated that 60% of the shipping accidents occur due to human error. Under the topics addressed within the center, special emphasis is given on path following and collision avoidance which are two important aspects of ship navigation. Collision risk assessment and avoidance is one of the major topics that has been covered under this proposal. Development of a robust algorithm for collision avoidance pursued under this will ensure safe ship transit in the future.
The centre will also look into other arenas for funding and sustenance. Both the sponsored research and industrial consultancy will be promoted under this center. From navigation autonomy to complete autonomy will be the future, and proposals will be submitted to national and international funding agencies in this area. The centre will submit proposals to funding agencies like DST, SERB, NRB (Naval Research Board), Ministry of shipping, office of Naval Research (ONR) for availing research funding in the field of ship autonomy.
The know-how in marine autonomy gained under the projects executed by the centre will have great values in both academia and industry. We foresee stakeholders from both industry and defence sector who will be interested in autonomous ships for inland navigation (ferries and boats), hydrographic survey (NIOT and NIO), search and surveillance (Indian navy and coast guard) and offshore platform support vessels (ONGC and Reliance industries ltd.) etc. Moreover, the ship designers and builders will be interested in conducting various types of seakeeping and manoeuvring (simple to complex type) and the centre will be able to provide service to them.