Digitalisation and Automation
in the Maritime Industry

At Data Respons Solutions, we are experiencing increased interests as nearly everyone is engaged in digitalisation and automation. We are actually observing the same mechanisms within ships and offshore as in land based transportation, says CTO Hans Christian Lønstad in Data Respons Solutions. The development has ripple effects for the entire transportation chain – with autonomous ships comes autonomous ports.

FLEET MANAGEMENT SOLUTION

We are currently working with a large transportation company to develop an advanced fleet management solution. – It is a mix of new and existing equipment, some retrofit, set up with sensors and 4G gateways, Lønstad explains.

AUTONOMOUS PORTS AND PORT LOGISTICS

Our customer also works with port logistics: – The plan is to deliver digitalised subsystems that contribute to increased automation in container ports.
It is essential to know where assets and machinery is located, and by using accurate (augmented) positioning systems. We are also working with a subcontractor on intelligent video solutions, based on machine learning.

Cranes and crane intelligence is an important part of port logistics and for customer Lehnert Cranes we have supported the development of a crane anti-sway, positioning and collision avoidance solution, with special purpose embedded computers and cameras.

–  It should be noted that innovative systems could be developed without AI. Fully automated ships may ascend by using augmented sensors like LIDAR, Proximity Sensors and IR cameras. These new sensors can for example assist in docking procedures where a high degree of precision is required, says Hans Christian Lønstad,  CTO of Data Respons Solutions.

CHANGED COMMUNICATION NEEDS

When it comes to technology development, 5G will play an important role in automated solutions, as the mobile network can be integrated with local solutions (Edge computing). In addition, if the servers are directly connected to the cells, the communication needs can be reduced by running local control loops.

HUGE AMOUNTS OF DATA

There are major changes going on in the more traditional maritime sector too. We have delivered many maritime-certified displays, panel PCs, etc., but also see that there are more camera-based data included in for example autonomous steering, he says. -It means huge amounts of data, and we are currently working with our partners to expand the capacity of the LAN ports, says COO in Data Respons Solutions, Jørn Toppe.

MACHINE LEARNING IS COMING

In addition to the relatively powerful, Intel-based, processors in maritime computers, typically NVIDIA processors or equivalent are capable of running AI processing as machine learning / neural network algorithms. – We are positioning ourselves for this development, emphasises Toppe, and when new requirements for AI comes, we will be able to build AI and machine learning into the systems.

–  It appears that capital is navigating towards sustainable, green technologies. Autonomous systems are certainly sustainable, more efficient, use less energy and are far more economical, says COO in Data Respons Solutions, Jørn Toppe

WARNS AGAINST HYPE

As an old cybernetics man from NTNU, Toppe is engaged and excited about the development but warns against the AI hype. – No doubt AI is a trendy word but it is difficult in practice. You need to train a system exceedingly before it is possible to create a sensible algorithm for the system. Still there will probably be holes in the training, which can have consequences, however, it is coming – that is for sure, he points out.

– There will be safety-critical functions that cannot use AI directly. It is however likely that AI will be able to gain great importance in industrial systems, says Lønstad. Especially promising is camera / machine learning used in sorting and quality assurance in connection with production.

AUTOMATION AND AI

It should be noted that innovative systems could be developed without AI. Fully automated ships may ascend by using augmented sensors like LIDAR, Proximity Sensors and IR cameras. These new sensors can for example assist in docking procedures where a high degree of precision is required.

Before throwing AI at everything, some fundamental issues concerning AI should be evaluated. The output of an autonomous system governed by machine learning is based on a learned probability distribution applied to the live sensor data. The effect is that specific decisions made by the system, becomes hard to evaluate and document in comparison to a direct coded behaviour that has undergone scrutiny by regulatory bodies. This will likely limit the proliferation of AI in systems where safety is at stake.

A CLEANER AND MORE SUSTAINABLE INDUSTRY

The global energy consumption will continue to drive investments towards a cleaner and more sustainable industry.

– Norway is fully capable of taking a leading role in autonomous, maritime systems, says Toppe. – It also appears that capital is navigating towards sustainable, green technologies. – Autonomous systems are certainly sustainable, more efficient, use less energy and are far more economical, concludes Toppe, who looks forward to Nor-Shipping trade fair, where Data Respons Solutions will be present to help current and future customers navigate the digital transition process.