Thomas Kumm, Global Technology Manager, Ports, at ABB Marine & Ports
Applications which use object recognition and sensor fusion are easier to industrialize at scale than ever before, and the results include more opportunities to automate container terminal operations and continuously enhance equipment performance. However, it is critical to distinguish between the general impact of AI across society and the opportunities it brings for port-specific activities.
Knowledge is AI power
I find the flexibility of AI to align with the container terminal's strategy to be one of its most compelling benefits. AI can be applied to increase levels of automation and drive container cranes towards autonomous operations. But it can also be deployed to assist crane operators with their workload, for example by improving graphical guidance at the user interface or enhancing safety through better risk detection and more responsive warnings. AI can add value to all port operations.
AI through machine learning is only optimized when trained with the right lessons.
ABB’s decade-long experience in developing and delivering crane automation systems for container terminals worldwide provides a strong platform to fully exploit AI. For example, it is the basis for a step-change in the power of ABB Optical Character Recognition (OCR) systems where AI enhances the visual recognition of container ID numbers and other features, with continuous ‘training’ as new images are captured. Over time, recognizing more features without exception management improves performance.
It’s my experience that simply adding external AI functions to crane automation will not optimize performance; instead, AI needs to be developed from the beginning to meet defined goals, then fully integrated, tested and verified in crane operations. Clearly, applying AI technology in ports is not as easily transferable as some solution providers would claim; rather, AI through machine learning is only optimized when trained with the right lessons.
With experts located at our R&D Center in Västerås, Sweden and our execution units around the world, we at ABB Ports work closely with customers all over the world to improve their applications using AI technology. We upload integrated data, perform training based on the data and download new algorithms for the customer to use in a continuous improvement process, making sure the right lessons are trained.
The AI ahead
The uniformity of container dimensions and the repetitive ways in which they are handled have made container terminals a natural fit for automation, with the first automated yard cranes entering service over 30 years ago. Today, yard operations can be highly automated, picking up and landing containers with minimal manual intervention, and remote supervision. We see AI technology used in the interaction between the yard crane and outside objects such as external chassis, to increase automation and reduce manual interventions even further.
Using AI, we can create a new layer of terminal safety, integrating visual and sensor-based data systems to distinguish humans from other objects in and around the crane, including new processes.
With accidents having affected attitudes towards autonomous vehicles, I want to emphasize that for ABB safety comes first. AI can already improve safety in container ports, enabling terminal operators to move work from under the crane to a safe office environment. With more autonomous functionality in container cranes, new safe solutions are under development and processes in the ports need to be adjusted for more autonomous cranes to ensure safety. This means new ways of working as well. The hands-on work in proximity of the cranes will be repurposed into new roles within maintenance and process supervision.
Our sector-specific experience indicates that AI-based solutions have reached a decisive moment in container terminal operations. In addition to focusing on practical solutions which improve safety and enhance efficiency today, we are convinced that our latest work will overcome barriers to further automation tomorrow and can even change the business model for container port technology.
It is important to remember that ‘success’ is not necessarily measured by the amount of automation AI delivers. Instead, it will be measured by AI’s contribution to the business of container handling and the needs of all stakeholders.
One example of a technology whose advancement interacts with AI to take these next steps is Light Detection and Ranging (LIDAR). It measures distances with precision using a laser to target surfaces and gauge the time it takes for reflected light to return to the receiver. As LIDAR becomes more affordable, it can be combined with higher definition digital cameras. Through sensor fusion systems, LIDAR can ‘see’ and recognize objects more effectively.
It is a well-known fact that automation in ship-to-shore operations has proved more challenging, like the added variable of ship movements in the water, as well as humans operating near the ship-to-shore cranes. Using AI, we can create a new layer of terminal safety, integrating visual and sensor-based data systems to distinguish humans from other objects in and around the crane, including new processes. Similar technologies are used in urban environments, however – as before – applying them to container terminals leans heavily on ABB’s operational experience.
It is also fair to acknowledge that opinions vary on the desirability of further automating the ship-shore crane move. It is important to remember that ‘success’ is not necessarily measured by the amount of automation AI delivers. Instead, it will be measured by AI’s contribution to the business of container handling and the needs of all stakeholders – from terminal managers, customers and third-party logistics providers, to crane operators, terminal equipment drivers, service engineers and stevedoring unions. All of these parties need to work together to agree and implement solutions that increase safety and productivity.
I find the flexibility of AI to align with the container terminal's strategy to be one of its most compelling benefits.
From every perspective, the huge amount of historical and current data ABB holds across multiple container terminal applications from ports worldwide is a key to successful AI implementation. This data is available to train and verify algorithms to interface with different ships, terminals, scenarios and exceptional events, ensuring that the algorithms are trained before an ABB solution is implemented at a customer site.
Toward smarter ports
Regardless of the attitudes towards AI, port stakeholders cannot ignore its built-in ability to enhance machinery performance. Traditionally, for example, container cranes have been delivered with a specified performance that will remain the same until the crane is updated after 10-15 years with new hardware and software.
AI and machine learning change this paradigm because they retrain the crane operation algorithms. Terminal software can be updated on a frequent basis, either to enhance efficiency, to enable greater automation, or to optimize on other parameters – according to the terminal operator’s preferred course.
AI can add value to all port operations.
AI also offers the opportunity for automated systems to develop an ever-increasing understanding of control parameters and other variables on the fly.
We are already seeing how AI technology opens completely new possibilities for container terminals, leading to safer and more productive operations. The use of digital tools shortens time to market and what we see today is only the beginning.
The rapid progress of AI combined with ABB’s deep automation expertise and industry knowledge will help us overcome barriers and find better solutions for the real-world environments, also considering new business models for container port technology. We are here for the change.
