The Manila International Container Terminal (MICT) has been implementing artificial intelligence (AI) at its container yard.
As part of the implementation of AI to draw out efficiencies the terminal operator worked with IGO Solutions and Avlino.
IGO Solutions has worked with automated terminals since 2006 while Avlino brought AI-expertise but had no previous experience in container terminals.
“Compared to the traditional AI platforms we came up with a reference architecture that has one unique capability or patent pending technology which is called feature discovery or feature engineering,” Ramana Jampala, President & CEO, at Avlino, explained to PTI. “Which means without having to know the underlying dynamics of an industry if you give the platform the underlying data sets that are required it can come back and give you ways to reduce RTG moves for example with changes to certain perimeters or your rehandles can be reduced if you change the business rules from the traditional system.”
The high-performance feature engineering with AI platform was brought to ICTSI in Manila and superimposed with the business cases that had been outlined.
Working with ICTSI, Jampala said the fact they were willing to share the last one to two year’s of data in order to train the AI from the outset was a real advantage.
The confluence of IGO Solution’s industry experience, Avlino’s AI technology and ICTSI’s participation has helped the development of an entire suite solution which Jampala said can realise the true value of automation.
In 2021 the company is implementing the YardSight AI module, part of the AiCON solution suite developed by Avlino. The overall solution suite covers everything from berthing to the gate.
One of the main parameters for the project in Manila was to minimise the rehandling and use all of the container handling equipment (CHE) equally. In addition, the AI was expected to make simultaneous decisions in order to find the right answer.
Change in approach
Jampala was keen to highlight that the approach was not to provide a platform with the onus on the end user to draw value from it.
He also differentiated from it being consultancy or business workflow software from which they are trying to do predictions.
Instead, the approach is to integrate comprehensive infrastructure and process into it and use an intelligent AI which can work on its own.
“Once you have done that,” he said, “you can do all of these things together, that’s when you can go ahead and say that I’m not going to tell you what should be done, rather I will move the cranes, I will tell you where the containers need to be.”
“If you are able to deliver that kind of a solution then you are genuinely providing a solution rather than a platform or a product.”