Getting ready for AI

In-Depth

Tideworks is developing a “strategic framework for establishing AI in terminal operations” that aims to put terminals in a position to take advantage of the new technology.

There is a fair amount of excitement from container terminals about how AI tools can help TOS modules better organise a container yard to make the process of exchanging containers between different transport modes through a large storage yard much more efficient.

Adding further excitement is the fact that investment in AI is not slowing down. Companies including SAP, Fujitsu, and Envision are now promoting their AI tools for container terminals.

“Peerless at analysing huge amounts of data, AI is aptly suited to help port terminal managers plan and orchestrate in real-time the many changeable variables involved in container management. These include scheduling people with the right skills and available equipment, such as cranes and automated vehicles, in sync with shipment arrivals and truck and railway transport. AI can quickly factor in special storage and handling rules for cargo types, as well as authorisation and other data from bill of lading and customs documents,” declared a recent article in Forbes.

In an interview with WorldCargo News Chad Van Derrick, vice president of software product management at Tideworks Technology, said its container terminal customers are understandably excited about the potential of AI.

Van Derrick sounds a caution, however, over expectations that AI will, by itself, be able to unlock a new level of efficiency in container terminal operations through data insights alone. The reality, he stressed, is that AI is very powerful but terminals need to position themselves to take advantage of the new tools.

Data quality

Tideworks see a direct line between AI and earlier technologies, including optimisation software and rules- based planning applications. “At the foundation is data,” Van Derrick said. Whether it is AI or another data technology, terminals need to understand that data-driven technology requires people, processes and technology that work together seamlessly.

At many terminals, there is a big gap between the data in the TOS and the actual operation of the terminal. This includes work instructions that are overridden by planners and equipment operators, and yard positions that are confirmed manually to the TOS when they are not actually accurate. Van Derrick emphasised that training an AI learning model with bad data will not produce good results.

Tideworks is helping terminals pre- pare for AI by developing a “frame- work for AI” that starts with identifying and closing gaps between people, processes and technology. Experience from implementing earlier optimisation tools has taught Tideworks that planners will override results that they cannot understand, or seem to be counterintuitive. Equipment operators will sometimes continue to make operational decisions on the fly.

AI will not change this by itself. To deliver value AI tools either need a fully automated environment, or the software has to be implemented in combination with processes that terminal personnel understand and work with.

Snowflakes

Van Derrick stressed that all terminals are slightly different. As such, Tideworks is developing and validating AI tools specifically suited for different use cases. He described possibilities that ranged from help desk chatbots to automatic yard optimisation algorithms.

Van Derrick declined to provide details at this point, but Tideworks is continually working on evaluating the latest technology for inclusion in its offerings, including software from some AI companies, to verify if, how, and in what circumstances, their tools can improve container terminal productivity. It is also working with AI leaders, including Microsoft and Amazon Web Services, to understand their approach and what this can bring to the terminal sector.

Tideworks firmly believes its measured approach is better than “taking a deep dive into the swimming pool” of AI offerings and working backwards to develop a viable business case for new ‘AI-based’ products. Van Derrick added that he is convinced AI is going to be a game changer in certain areas, but the very first hurdle remains at the port. Terminals need to close process gaps and implement the data tools so they have quality real-time data that AI can work with.

You just read one of our articles for free

To continue reading, subscribe to WorldCargo News

By subscribing you will have:

  • Access to all regular and exclusive content
  • Discount on selected events
  • Full access to the entire digital archive
  • 10x per year Digital Magazine

SUBSCRIBE or, if you are already a member Log In

 

Having problems logging in? Call +31(0)10 280 1000 or send an email to customerdesk@worldcargonews.com.