Main problem addressed:
Growth in international trade and the construction of larger container ships have had a major impact on road operations for the entry and exit of trucks to/from ports, sometimes causing significant levels of congestion, which in turn have led to high levels of pollution.
Although ports know the containers that are unloaded from a vessel, they do not know the real date of departure of those containers from their facilities since this date depends on whether these containers will be inspected by Customs, as well as on their final destination, days off without terminal storage costs, type of cargo (e.g. dangerous goods), among others.
Despite the cruciality of this information to help port managers with their planning processes, there is no AI tool that predicts the number of trucks that are going to come or depart from the port on a given day. This lack of valuable information means that appropriate measures cannot be taken to mitigate congestion and the levels of pollution and accidents caused by it.
This pilot case study intends to integrate different platforms, sensor networks and sources of information to predict the date and time of entry and departure of trucks using predictive analytics and business intelligence tools. In this way, and by achieving high accuracy in the predictions made, it will be possible to determine how many trucks/hour will leave and enter the port at a certain future date and time. With this knowledge, port managers will be able to anticipate possible congestion problems in the port, making smart decisions (e.g., enabling more exit/entry lanes, allocating more resources, spreading transport requests across a time period, carrying out operations such as massive transfers at a certain time of day, etc.) that will avoid congestion as well as the consequences that it generates: pollution increase, higher accident probability, and more.