During the World Port Hackathon in the harbour of Rotterdam on 4 and 5 September 2015, Finaps won the Highest Potential award for their application “The Danger Zone”. The award was for the application that has the highest potential of being marketed and implemented. Finaps honoured its name as Hackathon champions again!
The challenge Finaps addressed was to perform smart analysis on patterns and causalities of transporting dangerous cargo’s. Large amounts of datasets are already available however much is to be gained in smart analysis of this data in order to make meaningful predictions, moving from a responsive to predictive modus for the prevention of danger zones.
In 24 hours Finaps developed an application that can predict different levels of danger in a predefined area. Multiple data sets and API’s were integrated. These data sets are linked together based on geo locations. Using predictive models it is possible to provide insight in the segregations of hazardous cargo, events or other parameters to avoid the creation of danger zones.
By using deep learning algorithms, insights are gained into causalities that would not be evident using current methods. A rule engine was included in the solution to better capture explicit user knowledge. With this the years of experience from people in the port can be leveraged. When all this is set up the Danger Zone learning engine will continuously predict and/or visualize danger levels over a predefined area. The use cases for such predictive risk maps can be used in industry areas such as safety control, insurance rate calculations for ships and governmental policy writing. The possibilities are endless.