Estimation of Remote Operated Vehicle Motion in XY Plane using Unscented Kalman Filter

Indonesia, well known as a maritime country, has a very wide water territorial. Exploration carried out nowadays on average is limited to the sea surface while underwater exploration is still very rarely conducted. With technological advances in the marine and shipping fields, it is hoped that it can support the optimization of underwater surveys and exploration. The underwater surveys and explorations already conducted still used conventional methods, namely divers plunging directly into the water, then conduct surveys and explorations without the help of any tools. The obstacle that divers encountered with was the difficulty of the dive field resulting in an in optimal exploration. Not only did the underwater surveys require high technology, but also did the detection of drowning victims in reservoirs or lakes also requires diving equipment to find drowning victims. Given a certain depth, water pressure can endanger divers. The use of robotics in the shipping discipline is expected to assist in the detection of drowning victims. Robots commonly used to assist these explorations are underwater robots. ROV (Remotely Operated Vehicle) based water robots in developed countries have been used in underwater exploration. The ROV is considered more optimal in underwater exploration compared to the conventional methods. This paper is to apply the ROV motion estimation on the XY plane or on the water surface to help the SAR team find drowning victims. The estimation method used was the Unscented Kalman Filter method. The simulation results show that the UKF method could effectively estimate the position of the ROV on the water surface or the XY plane with an accuracy of about 98%.