"The reliability of shipping companies' schedules has dropped to 30%"

"The reliability of shipping companies' schedules has dropped to 30%"
Doron Broitman , 08.09.22 https://www.calcalistech.com/ctechnews/article/a609cyqu5

“Out of 10 cargoes, 8 we manage to predict better than the shipping companies," said Ami Daniel, CEO and co-founder of Windward

"As private consumers, when we order a product from Amazon we know exactly where it is, but when you order a container that costs $15,000 you don't know where it is," said Ami Daniel, CEO and co-founder of Windward, speaking at Calcalist's National Conference for Trade Infrastructure & Innovation at the Ashdod Port.

"The reliability of the shipping companies' schedules has dropped from 80% before the pandemic to 30-40% today. That means only about a third of the containers you ordered will arrive at the time they told you they would. There is a huge problem here. The collapse of the supply chains during the pandemic helped us understand this.”According to Daniel, "World trade has reached the $14 trillion mark, with $12 trillion of them passing by sea. This is not going to change for the simple reason that it is much cheaper to ship by sea. We can predict when a container will arrive much better than the shipping companies themselves. The reason for this is very simple, a shipping company manages its containers, but each cargo owner works with all the shipping companies - everyone works with everyone but only know about their specific cargo."

The technology developed at Windward is based on AI that processes and analyzes data from containers across the world. "Each load is scanned and that's how you know how to track it and understand where it is," explains Daniel. "We merge this information with the ship's data on where it is and what it is doing. The system checks between 60-70 million free parameters in several networks that analyze the information and makes predictions based on historical data. Another network checks the shipping price data and another network checks the cargo itself. The result is very simple: out of 10 loads of cargo we manage to predict 8 of them better than the shipping companies."