*Environmental Health*

conditions. Today the shipping industry is living its technological age. The most significant benefit of this age is "being available of data transfers from ship to shore". With these advantages, we are going to find the opportunity to develop autonomous ships.

While the shipping industry is talking about autonomous ships, the rest of the industry has already started to use robots and artificial intelligence in the industrial activities. However, before the autonomous ships get into forced, we should find answers to these questions:


In the current study, an empiric data mining and machine learning was applied to the real sample data from a vessel which is given in **Tables 2** and **3**.


#### **Table 2.**

*Extracted Voyage Data From Table 5.*

*A Review of Alternative Marine Fuels DOI: http://dx.doi.org/10.5772/intechopen.97871*


**Table 3.**

*Accuracy rate.*

When we look at the literature on data science methodology, we come across with different kind of analytics. Autonomous ships must use descriptive analytics that recognise the data, predict the data based on the description, then prescript the data and take action according to the traffic congestion found and predict the current condition from the history of the data. Prescriptive Analytic uses the results of the descriptive and predictive analytics. While descriptive analytics are evaluating the current data, Prescriptive Analytics examine the data and gives suggestion and takes the actions without a human. All the prescriptive systems are managed and run by machine without a human.

As an empirical application, the author used the data given in **Table 5** in order to predict the data shown in **Table 4**. The data in **Table 5** represents the real ship data collected from a voyage between Port of Norfolk to Port of El Dekheila during authors previous work experience with a largest shipping company in Turkey.

After performing following six-steps, the data presented in **Table 3** was achieved.

1.Excel reader

2. Statistics

3. Scatter Plot


With the scorer node, the author checked the accuracy of the learner by prediction results.

Depending on the data's properties, the accuracy rate has been changing. In this data sets, the learner can predict the results with %33 accuracy. Healthier data and different partitioning tools can decrease this rate.

As we understand from the tree in **Figure 8** when the swell direction is from "N or north", the slip is going to be more than %14,5 which means that the consumption of the fuel oil and greenhouse gas emissions are going to increase.

By the help of this simple prediction model, the company can easily predict the engine slips from the up-to-date data getting from the ships. The distance of the


**Table 4.** *Predicted Slips.*


#### *Environmental Health*


#### **Table 5.** *Navigation Data from Norfolk to El Dekheila.*
