The first prototype of a fully autonomous ship could be a reality by 2015, and they could be commonplace by 2035, according to recent research conducted by DNV GL. The organisation has developed the concept of the 'connected ship' to describe the convergence of real time data transmission, high computing capacity, mathematical modelling, remote control, sensors and miniaturization that could change the shape of the shipping industry.
These new technologies could play a key role in reducing shipping fatalities. “Online monitoring and decision support can lift human performance, closing the gap between safety goals and current practice. Shipping currently has a crew fatality rate 10 times that of OECD best practice levels, and the industry needs to target a reduction of 90% to match that level,” says Dr Tor Svensen, CEO of DNV GL Maritime.
ICT solutions can provide control over the status of degradable systems, increase situational awareness and human reliability, support in the definition of corrective actions and reduce operational risk. In realising this potential, shipping will benefit from developments in the offshore, aviation, aerospace, and automotive industries which have been the primary drivers for advances in automation and remote operations.
“Shipping will likely apply these technologies to instrumented machinery first and then gradually to vessel navigation, which will be operated remotely from shore-based centres,” says Dr Knut Erik Knutsen, senior researcher at DNV GL. “These solutions will increasingly rely on sensor technologies and computers to manage onboard systems from remote locations. As more systems become automated, the number of onboard personnel will be reduced, and more decisions will be made from shore-based control centres.”
Onshore control centres will be responsible for the condition management of the ship and risk related to the failure of onboard equipment or broken communication links.
They could eventually be responsible for operating vessels in congested sea lanes, or in proximity to ports and terminals and in emergency situations. To manage these tasks, control centres will be equipped with system simulators designed to select optimal routing procedures and interfaces with land-based supply chain networks. Such systems should provide accurate representations of risk and allow humans to take full control of vessels from a remote location when necessary.
“The correct adoption of the new technologies that could make this happen requires a big change in the culture of shipping companies, including the development of trust in the black box,” says Dr Knutsen. This, as a first step through smart maintenance, is the ability to rely on increasingly sophisticated systems that can indicate a developing failure.
Smart sensor networks will be critical, as their ability to work together offers a detailed and accurate picture of complex onboard systems. The sensors will be able to react to changes in their surrounding environments and reconfigure themselves in order to perform multiple functions.
Fast-forward a few years, says Dr Knutsen, and tasks will gradually be handed over to computers that make the decisions regarding performing maintenance and allocating resources. Systems will be semi-autonomous in the sense that they will be able to decide on maintenance needs with limited inputs from a human operator.
“For shipowners of today, it is important to consider the stages of implementation of a highly advanced condition monitoring system. Benefits will also come in stages, depending on the level of commitment in terms of corporate cultural changes and willingness to invest in development and infrastructure,” he says.
The first stage of implementation is to determine the most critical failure modes of a ship system and to equip the relevant components with sensors that can detect related physical parameters. This information is analysed by onboard computers that perform diagnostics and determine the health state of the component and whether it should be inspected, overhauled, or discarded. The direct benefits of this first stage are the capability to prevent failures and reduce costs and safety concerns.
The second stage includes the use of prognostics to determine the remaining useful life of components and to gain an understanding of when maintenance should be performed and to plan for that. Prognostics use statistical information about failures or physical degradation models, together with continuously measured parameters that determine health state, to determine remaining useful life. The benefit of having access to this information is that the reliability of the ship is estimated in real time, and actions can be taken if needed.
“As experience with such systems grow, and the related mathematical knowledge of failure processes develops, the accuracy of both diagnostic and prognostic algorithms will improve and give more accurate residual life estimates. This will allow further switching towards online, risk-based maintenance, where maintenance is seen as the set of actions that is used to maintain or reduce the system risk level cost-effectively. Here, the difference from the old view of risk-based methods lies in the fact that these measures are evaluated real time and from actual operations, rather than being based on recommendations and poor statistics,” says Dr Knutsen.
The ultimate goal for Dr Svensen is to be able to deploy integrated systems, advanced navigation systems and sophisticated software that could manage smart sensor and actuator networks to maintain a vessel’s course in changing sea and weather conditions, avoid collisions, and operate the ship efficiently. Shipping could draw lessons from the offshore industry, which has developed a number of automated systems with marine applications to improve performance, he suggests.
Remote operations will require automation of the engine and other integrated systems, alongside advanced navigation systems and sophisticated software that can operate the ship efficiently, within specified safety parameters. The onboard ship control and decision management system could be adjusted to allow different levels of autonomy, but with further advances in these enabling technologies, Mr Svensen envisages a completely autonomous ship that reports to shore-based operators only when human input is needed or if emergency situations arise.
Combined with advances in materials requiring limited maintenance, autonomous shipping would eliminate occupational risks onboard. While the unmanned vessel concept would likely face significant public scepticism, DNV GL believes that the use of many autonomous systems on board ships will improve the industry’s safety performance.