Trucks arriving at terminal gates in an uncoordinated manner, difficulties in manning the crew on board in the right number and with the required qualifications, ships lying at achorage after weeks of sailing – in the maritime trans- port chain, frictional losses and inefficiencies can lead to avoidable costs and reduced productivity. In many cases, a targeted analysis of existing information reveals weak- nesses and optimization potential. We at Frauhofer CML accept this challenge and de- velop solutions for practical applications.
Maritime companies accumulate digital data in many different forms and formats in their business activities
– sometimes systematically, sometimes unintentionally. The data come from various sources, for example from ship sensors or fleet management systems, and additionally navigational and technical operation data are available. Much of this data often lies unnoticed and scattered on the servers, but it can be used to gain capital for optimizing further operations. After all, the correctly assembled com- bination of this data provides information that can form an important basis for future decisions.
Customized data evaluation The Fraunhofer CML has the competence and the methodology to analyze and in- terprete data in a targeted manner. Because even though much can be done with algorithms in data analysis, the art lies in knowing how to prepare the data. This requires process knowledge. By recognizing similarities and patterns in data sets, for example, an unmanageable database can be categorized and made accessible.
However, apart from the question of what data is available or how to access it, entrepreneurs often have no concrete idea of the benefits that data analysis can bring. Especially in maritime logistics there are many areas of application.
Optimized truck handling Data can be used, for example, to better forecast truck arrival times and thus improve traffic flow in the port. For this purpose, the Fraunhofer CML developed a model that uses a digital image of the handling processes of logistics nodes such as port terminals to achieve optimized handling by predicting truck arrivals. This method uses historical and current data and is based on an artificial neural net- work, which can take into ac- count further influencing fac- tors in the form of so-called predicted values. This can reduce planning uncertain- ties and achieve optimal truck scheduling for terminals, for- warders and truckers, which reduces avoidable costs.
Flexible crew planning on board
In another project, the soft- ware solution SCEDAS® was developed to plan the deployment of personnel on board a ship efficiently and in accordance with legislative regulation as well as company specific rules, using mathematical optimization methods. In addition to the special demands of a specific voyage on the crew and their qualifications, SCEDAS® takes into account legal requirements and thus supports the complex task of crew management on land and on board. In the mean- time, the SCEDAS® crewing software has been further developed so that maintenance and service tasks are integrated optimally in the work schedule.
Safe and efficient sea voyages
The analysis of data from the Automatic Identification System (AIS), which among other parameters transmits position, speed and course data of ships at sea, enables route optimization and the forecast of ship arrivals. Based on historical data (AIS has been used by all merchant ships since 2002), optimal voyages can be deter- mined, but also critical sections with heavy traffic can be identified where increased attention by nautical officers is required. The correlation of AIS data with weather data allows improved up-to-date route optimization, which can significantly increase the safe- ty and efficiency of a voyage.
Talk to us!
These are just a few examples from the maritime industry where data analysis has been able to create added value for our customers. In further pro- jects the evaluation is also al- ways user-oriented and driven by the question: How can the analysis help the customer to optimize his decisions, or how can he use his data in a meaningful way?