Predicting Fuel Consumption

By leveraging existing sensor data it was possible to predict fouling levels on ships. By predicting when the fouling starts being a highly negative factor, it is possible to optimize the fuel consumption which accounts for 80% of OPEX.

The problem

J. Lauritzen has many vessels geographically distributed globally. Vessels consume enormous amounts of fuel. Hence fuel is one of the main cost drivers in the industry. Fouling increases the oil consumption significantly, however, fouling slowly increases over time making the increasing impact almost impossible to detect. Fouling is typically dealt with by visual inspection underneath the vessel which is impractical and expensive. Thus, often the amount of fouling is unknown to the vessel operator.

Increased water resistance, due to algae on the vessel (Fouling) is a huge cost-driver withing the shipping industry. By predicting when fouling is a problem, the required actions can be carried out to remove the fouling.

Our solution

We were hired by the vendor (Delegate) to build a data driven support tool for their vessel operators. We trained an AI model using data gathered from 100+ data sensors located on the vessels. The model recognizes data patterns indicating an increasing amount of fouling on the vessel, enabling the model to identify equal increases in real time on operating vessels.
Results were presented in Microsoft PowerBI providing the vessel operator with high-quality information on the fouling state of the vessels. Vessels can then be cleaned whenever fouling becomes significant ultimately reducing the overall fuel consumption.

Other cases

We use our RPA platform to extract data and in an automated process, we conveniently provide you the insights you need to stay ahead of competitors.
In order to mitigate potential losses in regards to project budgeting and unanticipated expenses, we investigated the use of advanced analytics as a tool to help Project Managers with budget control.