The ship and the new supply chain
Autonomous ships “attack the conventional wisdom on economy of scale”, says Ørnulf Jan Rødseth, senior research scientist at Norway’ independent research organisation, SINTEF.
“The big 20,000 teu ships are extremely efficient,” he says – that is, if you keep your focus limited. “Unfortunately you end up with a few ports investing hard to take these larger vessels, which pushes port costs up and, overall, the result is a less flexible transport system, more reliant on transhipment.”
But, according to Rødseth, if you can bring down ship sizes, making them cheaper to run and favouring direct, point-to-point calls, “then I believe you can reverse the argument” and look for cost-effective routes connecting smaller facilities.
What he outlines is a highly-tailored operation: the very first, Yara Birkeland (which has just begun construction at Vard Brevik), will be more like a conveyor belt, with fully automated cargo handling and mooring on both ends of the route. As such, Yara shouldn’t be considered a ‘ship’ at all in the traditional sense “but a component in a dedicated cargo transport chain”.
It might have been slow to start, but it’s not ending with Yara. Rødseth’s colleague Håvard Nordahl, also part of the SINTEF-lead ASTAT project, explains that slow-paced electric cargo shuttles are being considered for the Jøsnøya, Trondheim and Verrabotn regions. Direct transfer of lumber from truck to ship, for example, could avoid multiple manual loading and unloading operations and even storage as one ship will always be at berth, ready to receive: “Day or night, it’s all the same,” points out Nordahl.
As these first projects may be carving out the pathway for future development, it’s worth noting that Rødseth says Yara’s particular control strategy is not just one way, “but ‘the’ way” forward.
It’s a deceptively simple idea: while the vessel has benefited from a lot of automatic collision avoidance development and the latest sensor technology plus sophisticated robotics, it’s still easier to plan ahead than solve problems on the hoof. “The operators will report to the VTS before entering the area, and that way it will avoid encountering the bigger ships,” he explains.
It requires close communication between the VTS and Yara’s coordination centre, but for the most part it relies on well-established technology. “Yara will always have people supervising the voyage, and that makes for a system that doesn’t have to be able to respond to every potential encounter alone,” he explains. Of course, he admits, it’s only possible because Yara is limited to a 6km run.
So, what about vessels on longer routes? It’s all about evaluating each distinct arena separately, and creating specific solutions, says Rødseth. After all, in the open sea operational challenges don’t tend to arise (or need resolving) quite as quickly, so in that environment, autonomous ships could potentially rely on satellite links to a coordination centre. On the other hand, for nearshore or inland vessels, a 4G or 5G network could eventually be stretched over reasonably significant distances.
“I even think we could change the way ships, all kinds of ships, sail”, says Rødseth as new developments will filter through to other, more traditional segments. Norway, supported by a very proactive coastal administration, has been developing the idea of ‘intelligent’ VTS for some time and sees the two elements, smart ships and smart fairways, working hand-in-hand, rolling out into other regions. There is potential for these interactive VTS routes “to transform the way traffic is organised”, he concludes.
CONTROL STRATEGY ‘STEPPING STONE’
There are other concepts, shaped by an environment that’s very different to Norway’s.
In Kiel, Germany, the authorities are aiming to decarbonise their entire transport and commuter chain, and Professor Thomas Meurer of Kiel University is leading a team that’s looking at creating control-ware able to launch, berth, and direct a zero-emission ferry across the bay.
This won’t have the luxury of wide open spaces: Kiel, says team member Max Lutz, sees particularly intense vessel movement: beside commercial ships, there are fishing, sail and motorboats “and although the bigger ships have some fixed routes there’s no real traffic separation scheme”. Further, he points out “the ferry probably won’t have right of way most of the time”.
While Meurer explains that “sometime in the future we will have a complete communication network around the bay”, as yet it doesn’t exist, so given the nature of the traffic, the intelligence has to remain firmly onboard.
Of course, this entails sophisticated decision-making architecture, “requiring active path planning” says Meurer. The development brings together sensors, actuators and mathematical vessel models to control the response, taking into account obstacle motion and collision avoidance. Part of the trick is to keep the calculations efficient enough to be able to predict inside a few milliseconds how the other vessels – big or small – will move and take appropriate action.
One interesting point is that the multiple inputs necessary for autonomous operation create not just complexity, but strength. Max Lutz says: “The classic example is GPS, gyro acceleration sensors and ship model. All these sources are somewhat distorted: the acceleration sensors can drift but the GPS tells you where you are. Then if the GPS jerks a little, you know from the model that your ship can’t just jump 5m to the left. So together they can make up for each other’s weaknesses.” But, he underlines, “it has to be done in the right way”, or the output could magnify, rather than diminish, the distortion.
While at the moment it’s still firmly a university research project, next year should see tank tests, and after that, a small USV in the Kiel Canal: however, getting to a marketable product will take time, “and a cross-industry partnership”, says Meurer. The ferry itself will probably be some years off, he admits.
Despite this, the development could still provide an intermediate stepping stone toward autonomy for a broad range of shipping operations. Its inherent flexibility allows it to be tailored to balance, for example, the fastest possible service with the most energy efficient “combining multiple objectives to end up with the optimal control profile”, he explains.
Therefore, the system could be of use to a wide range of ship operators, from fleet owners conscious of fuel and emissions, or those looking for a semi-automated point-to-point route with the benefit of obstacle avoidance. “In the end,” says Meurer, “We will have an automation system, from algorithm-based software, that can be integrated into an autopilot.”
Autonomy requires a complete upending of design approach, underlines Kourosh Koushan of SINTEF.
Certainly, an autonomous vessel will, at least initially, be more expensive than a traditional ship to construct needing an extra layer of redundancy on big-ticket items like power and propulsion – which Koushan points out, will be founded on electric and clean hybrid technologies.
Given the costs, advantages need to be pursued: “No crew and no passengers onboard firstly means a lighter vessel, a better range for the same battery capacity, and no energy consumption for the crew... It also means we can start thinking differently, there’s no one to get seasick, so we can allow different kinds of acceleration and motion.”
Still, there are limiting factors: effects like slamming and broaching will throw off the carefully constructed algorithms: “We need to design the ship in a way that minimises any erratic phenomena,” he explains. Likewise, the control systems also require stable propeller performance, avoiding pulsing, vortices and especially air ventilation “which degrades efficiency extremely fast”.
Even the hull itself will need more strengthening and thrusters protecting in case there’s contact with floating debris that the sensors have missed. Further, as part of an integrated system “we also need to think about the interaction with an automated dock”, he adds.
However, while ships need to be specifically tailored to the mission, his colleague Nordahl points out that reducing costs means “mass production... and standardising the systems, even ship designs”. The only way both demands can be satisfied is through a modular approach.
Interestingly, given that the business case is critical and rests on pinpointing the right solution from thousands of alternatives and moreover, there’s no underpinning experience in the field, Kourosh says that AI might play a part, breaking free of human-bias constraints. Artificial intelligence could present us with “radically new designs, which would not have been considered by a human”, he concludes.
LATEST PRESS RELEASES
Date: March - April 2019 Customer Name: UK Cruise Operator AtZ Contact: Project Engineer Clive ... Read more
When a customer in Singapore contacted Vickers Oils needing urgent help, we were able to deliver des... Read more
Trinity House, a UK based charity dedicated to safeguarding shipping and seafarers since 1514, has b... Read more
Stolt Tankers and DESMI Ocean Guard sign Frame Agreement for the Supply of CompactClean Ballast Wate... Read more
DESMI Ocean Guard’s CompactClean Ballast Water Management System receives amended US Coast Guard Typ... Read more
DESMI has gained a solid reputation over the years for selecting exactly the right materials for its... Read more