Autonomous Mobile Robot Localization using Markov Decision Algorithm Glenn Hanaya Sitompul (a*), Muhamad Rausyan Fikri (b), Ignatius Budi Susanto Hadisujoto (c), Iwan Setiawan (d)
a.) Mechanical engineering, Sampoerna University
b.) Information systems, Sampoerna University
c.) Mechanical engineering, Sampoerna University
d.) Information systems, Sampoerna University
Abstract
Localization is one of the fundamental problems of autonomous robot. This problem is basically related with determining the position of a mobile robot from sensor data. However, mostly solution to localization problems is passive, meaning, they do not manage the opportunity of controlling robot^s effectors during localization. This paper proposes an active localization approach, which is Markov localization. Markov localization provides rational criteria for setting the robot^s motion direction (exploration) and determining the pointing direction of the sensors to localize the robot most efficiently. Furthermore, it can deal with noisy sensors and approximative world models, which is far more beneficial. The data of our approach demonstration is taken using a mobile robot in a structured environment- therefore, it is assured.
