| Resumo: | Nowadays robots are becoming more present in our daily life performing a variety of on-demand services. In order to perform autonomous tasks the robot should be aware of its environment. To achieve this goal, there are three main problems to solve: mapping, localisation and navigation. During this work, we developed an autonomous welcome robot for the Institute of Electronics and Informatics Engineering of Aveiro ( IEETA ) with the capacity to receive requests from a visitor and guide him to the requested destination. At the end of this task, the robot should return autonomously to its docking station. To accomplish this goal we studied algorithms related to the three referred problems. As an example, a laser-based solution is used for the Simultaneous Localisation and Mapping procedure ( Gmapping ), the adaptive Monte Carlo localisation approach (AMCL) for the robot moving in 2-D and A* as a method for path planning. Improvements have been made regarding the use of these algorithms including in the environment an active localisation system based on the use of ultrasound beacons. The end result is an autonomous agent capable of mapping the building, self-localise in the resulting map and moving from current position to a specified target. It is also capable of path recalculation and minimal real-time collision avoidance while navigating.
|
|---|