Each year 9 billion tons of goods, representing more than 80% of globally exchanged goods are transported by sea.
Container placement optimization is today an important challenge for freight shipping companies. The ultimate goal is to make fewer container shiftings to save money.
Today, these companies use data terminal operating system, maintenance equipment system and equipment fleet which help them with container shifting operations. However, with modern technologies as Big Data and Machine Learning, we can do much better.
Every aspect of daily operations produces data, past shiftings, past, present and future orders, number of ships, capacity, maintenance operations etc… All these data can be taken into account in order to predict perfectly the future and optimize container placement.
More data can be also collected and taken into account in our prediction models:
The ultimate goal of a big data solution is finally to connect together all the data in order to correlate the impact of each of these data on container movement.
More data that we take into account when calculating a container movement prediction model :
For example, in France, the following day of a national football match will definitely impact the whole country’s productivity. Another example, the Ramadan will impact one-month activity.
In some regions, freight ships can cross difficult zones and have to take precautions before their load and unload operations. Political monitoring is important in container shifting.
Some ports are quicker than others for loading and unloading containers. Some are more expensive. Moreover, the free place within ports changes constantly. Therefore, in order to optimize container placement, all this information have to be taken into account.
Sea freights changes according to the supply & demand and also in function of the location. Many companies do not have the time to decide which cost will be the lower, they choose the easiest way.
When it concerns thousands of containers, computers are better than humans to predict which containers should be full and which one half empty. Nowadays many containers are only half full when they are shifting. With a continuous monitoring, it will be possible to fill fully containers before they even arrived at the harbor.
The aim of Big Data Algorithms is to mix all the data together, including terminal operating system, cultural events and all the other available data.
If Predictive Algorithms are able to put all the data together in order to predict the best ways for a container to reach a place and predict new orders, they will also be able to predict the best container placement strategies at a level that had never been reached before.