Proximo Puerto: Back to CTAC 2023, Automation & A.I.

I got back to CTAC (Container Terminal Automation Conference), 6 years later - still remember the last one in London where I iterated in the Automated Intelligence idea framed with Rich Ceci for NavisWorld 2017 and that we translated into an article for Port Technology International - Link to 2017 paper. https://www.porttechnology.org/technical-papers/automated_intelligence/

• ?The ideas at that already 'vintage' article continue being relevant, the belief is intact but the technology certainly progressed to make it tangible and this conference also helped to connect the Ideas where A.I. can help Automation to be easier and more repeatable.

For us the first one as NextPort - first of all Thanks to Port Technology International, to all the sponsors that made CTAC23 possible: and special appreciation to Benjamin Heather and Jack Donnely who along with the Conference Team did an amazing job, once again, organizing and hosting another reference event for our Industry:

• THANKS FOR THIS FORUM WHERE WE CAN DISCUSS ABOUT & PROGRESS ON THE WAY FORWARD.?

A.I. for Easing Automation

?As part of our Intro I got back to some of my first drawings in Hamburg in 2008, using Google Earth and trying to understand which for me still is one the most interesting references in Automation: Alterwerden by HHLA – a complex system of sub-systems that need to be synchronized.

And also reference while in Navis realizing the complexity around Screens, Information, Monitoring, Actions…: difficult and highlighting challenges for User Experience & Working Environments at Control Room in Automated Terminals.

• And my learning highlighted at the slide, connected to what A.I. could do for Automation: the most sophisticated algorithm won't make a difference.

?So with this in mind, what is for A.I. when helping Automation - we received good feedback:

When talking about Use Cases we put upfront a strong sense of realism & pragmatism around the role of A.I. to help dealing with changes but also with the scale of multidimensional decisions.

The whole Industry is asking Terminals to be more predictable: but How is not trivial, and A.I. for now is a small part of the solution…here some questions to answers:

?A.I. and Decision Making Evolution

?After the Intro we got into explaining the connection of A.I. with decision making for Container Terminals… A.I. is coming with the capabilities we need, on its different domains: forecast, diagnose, detection, segmentation, recommendations, learning…

• So with A.I. we can get our Terminals to be more intelligent, for sure also connecting energy & decarbonization goals as we discussed in Greentech here in Hamburg, some weeks back.

But it is not the holy grail – it needs to be understood, in terms of What it does, but also How can be applied (and the pre-requisites for this technology, which is NOT new at all but disruptive with the way we work today):

• In that direction we referred to the permanent schematic where we conceptualize OPS decision making across time horizons and across operational areas (decision domains):

?

• At the ‘look-ahead’ time horizon, the Planning Phase, the planners are sequencing, organizing, decking or loading decisions that are subject to change - they use the TOS but also additional tools like Berthing Modules, Yard Simulations, Gate Prediction Dashboards…and on-top manual or mental calculations based on their experience:

At Planning, the opportunity with A.I. is around generating visibility on the ‘look-ahead’, predictive capabilities that are not instrumented normally in the systems Planners have.

• At the Execution Phase, containers are discharged & loaded and when an exception is happening, the dispatchers or controllers have a hard time to figure out what is the right action to apply: as the root-cause of any exception is NOT trivial and sometimes taking a decision is ‘scary’ (You fix something and You can break something else).

?At Execution, the opportunity with A.I. is enabling a prescriptive approach at planning stage to avoid some exceptions: but also when happening, generating a system intelligence for users to anticipate problems and to be precise in solving them.

?No Data + No Interoperability = Impossible A.I.

?One of the most interesting outcomes from CTAC was the emphasis on standardization, in particular around Terminal Industry Committee 4.0 (TIC4.0) - the importance of having an underneath data foundation to support Automation in its different fronts (Equipment, Process , Decision), and beyond on A.I. A core foundation starting with:

1. Process Digitalization
2. Well-defined & Persistent Data?
3. Systems Interoperability?

?

A.I. value proposition is a complete hype if we don't have the core foundation on Data needed to Learn, Predict and Prescribe…

Then of course we need People at to embrace the A.I. practice, which is not trivial either. On the first two aspects we do have TIC 4.0 >> in my presentationI wanted to show appreciation for people like Frank Kho, Boris Wenzel, Norbert Kertner and of course Super-Curro Blanquer:

• while there still lot of work ahead, the generated potential and the perseverance TIC4.0 founder demonstrated it is quite amazing: creating a Forum to openly discuss and to progress in tangible steps to Terminal Digitalization:

?Digitalization is a pre-requisite for Terminal Automation - in coming years we will see its benefits to ease Automation and again to make it more repeatable.

?A "humble" start with NextPort for Terminals

?Finally we utilized the opportunity of being in CTAC to present our Product Ideation outline, where still in conceptualization phase we see the opportunity to complement the existing eco-system at Terminals (TOS, ECS, IOT…):

• This Team is 'here' for a while > in the journey to Data-Driven Decision Making the start is indeed putting the Data in Data Structures that can be Smart, meaning where A.I. can be applied.

Then the capabilities we are prototyping are simple, and fusioning two technologies that will help evolution in decision making at Terminals & Ports:

?The continuous use of a Process Digital Twin is fundamental to put Data in context, and as Dashboard, Prediction or Alerts intuitive to the User - generating simplicity for understanding the cargo handling processes, confronting 'What's going on' with enhanced a holistic & monitoring capability across systems.

• Then a continuous analysis capability with A.I., ML or Simulation to generate Predictive Capabilities for including ‘What-If-Scenario’ Analysis to generate those feedback loops mentioned before when connecting decision making at Planning & Execution stages.

Finally some guiding principles from our plan of approach for NextPort Terminal Product:

1. Working Backwards - very consolidated practice, and being Industry Veterans we know the Why, the What and the How for a solution like the complexity underneath requires continuous discovery & challenging, driven by use cases,??
2. Journey Partners - this Industry doesn't have enough collaboration. To go beyond we will need partners, but while exploring and being very open to partner, being selective to match puzzle pieces will be fundamental.
3. Ground Cable for GTM - Innovation is about learning when to say "No": we are extremely humble and pragmatic in our approach to Go-To-Market. A.I. so far has a high degree of exaggeration and there are simple problems still to be solved.

With this in mind we concluded our presentation with our Terminal Product Outline one-pager: we are looking forward for presenting in next CTAC how far we got in this 'journey' - Seguimos.