Localisation and planning of trajectories
Calculating the spatial location of a mobile autonomous robot, its positioning and orientation at each moment of time is one of the essential requirements for its navigation. Together with this, the capacity to plan movements that enable displacement from one place to another without colliding with objects or persons in the surroundings is fundamental.
The location of mobile autonomous robots involves determining the position of the robot in relation to a given map of its surroundings. The problem lies in the fact that its position cannot be measured directly and has to be inferred from sensorial data. Moreover, just one measurement is usually insufficient and the robot has to incorporate data over time for its position to be determined.
Not all location problems have the same level of difficulty. Depending on the initial knowledge of the robot about its position, two types of problems in ascending order of difficulty are differentiated:
- Monitoring position. It is assumed that the initial position of the robot is known.
- Global localisation. The robot does not know its initial position within its surroundings.
Clearly, localisation in dynamic environments is more difficult that in static ones. Most real environments are dynamic, with the presence of persons, moveable furniture and doors.
The planning of trajectories, on the other hand, involves finding a passable or negotiable path between a position of origin and a target position. The planning requires, amongst other things, a representation of the robot and its environment(s). The explicit representation of a wide number of these means that the computational costs of planning and memory are high, to the extent that it is unviable in a reasonable space of time. As a result of this, methodologies and algorithms have been drawn up to provide a compromise solution between one of an optimum nature and one incorporating computational costs.
At IK4-TEKNIKER probabilistic techniques are investigated for the resolution of the problem of global localisation, both in dynamic environments and static ones. The goal is to take the uncertainty into account and to enhance the precision of spatial location, applying various path planners depending on the characteristics of the robot and its surroundings.