Researchers at the Universidad Carlos III de Madrid (UC3M) in Spain have designed a real application for maritime surveillance that is able to integrate and unify the information from different types of sensors and data in context through artificial intelligence and data fusion techniques.
The system was designed by scientists from the university for Nucleo CC, a company that develops surveillance systems for the maritime and aeronautic sectors. The first prototype will be used in the near future in Cape Verde (Africa). Two types of sensors have been deployed there: a set of radars and a series of AIS (Automatic Identification Systems), which allow ships to communicate their position and give other relevant data on their location and characteristics. These two types of sensors offer complementary data, which can be fused in order to obtain better information as to what is happening in the maritime and coastal space of the area of interest. The system was created by scientists from the Applied Artificial Intelligence Group (GIAA) of UC3M, under a research project titled "Fusion de Informacion en Trafico Maritimo" (Information Fusion in Marine Traffic).
The result of this research, presented at the last International Conference on Information Fusion in Edinburgh, Scotland, has been the creation of data fusion software that allows improved maritime surveillance to be carried out, simultaneously integrating the capabilities of the radars and the AIS localization stations deployed. The objective is to guarantee security in an area by monitoring different ships that are in a given maritime route which, at the same time, is the entrance/exit of a commercial port. Jesus Garcia, one of the heads of the study from the UC3M Department of Information Technology, said, "It is necessary to have a complete, accurate, and up to date picture, similar to that which is provided to air traffic controllers, of all the ships that are in the area of coverage to be able to adequately manage maritime traffic and to detect anomalies as much in advance as possible."
The scientists developed a prototype that has been integrated into Nucleo CC’s system, after having undergone validation tests to be able to execute in real time with the data supplied by its sensors. It able to monitor 2,000 identifiable objectives between large and small vessels, with a capacity to process the data of up to 10 sensors, and provide the exit with one second refresh time. Location data for ships must be available "24 hours a day, 7 days a week, 365 days a year, independent of failures in the sensors or in the different intermediate mechanisms, and in some way, what this system attempts to do is guarantee that this can be done," explained Jose Luis Guerrero, another of the GIAA group researchers who worked on the project. "We are able to make it so these vessels never lose their position, thus avoiding collisions or any type of problem in information management regarding the movement dynamics of these ships," he added.
The scientists are now researching how to apply this information fusion technology to fields such as robotics, in unmanned vehicle navigation, artificial vision, or environmental intelligence systems. "In all of these areas," Professor Jesus Garcia said, "information fusion technology and infrastructure are necessary to combine the data from available sensors and the contextual information in each scenario."