DATA MANAGEMENT FOR SMART SHIP OR HOW TO REDUCE MACHINE LEARNING COST IN IoS APPLICATIONS

摘要

Shipbuilding process, generates a lot of information and data, which a priori makes it seem impossible to have all this data in real time, but the new processors, simpler and smaller, with a good connection to the Internet, make it possible. The data management is, however, only one side of the coin of the Internet of Ships (IoS). Energy efficiency is a fundamental aspect also in new devices that connect to the network. But IoS not only covers the stages of design or production of the boat. Once the sensors are in the components whose information wants to monitor, we will be able to obtain information throughout the life of the ship. IoS is presented as a solution capable of detecting when a component on a boat is close to fail and must be replace, when we take the boat to repair when we have to paint again, when corrosion has reached a certain limit ... and all this from our pocket tool and early enough to avoid late or unforeseen performances. IoS reaches this sector to ensure profitable production, or safe, efficient and sustainable process for all types of fishing vessels, tugboats, tankers, charges, ferries, dredgers and oceanographic ... Data management in Smart Ships, including collect, process, saving and third party distribution should be regulated, controlled and done in the most efficient and secure way for the ship owner. This can be accomplished based on a deep study of each part which is involved in close-to-real response systems and its machine-learning control unit. In these Systems, each parameter change generates stress and material fatigue, reducing its lifecycle.