The basic idea behind many mobile robot localization techniques is to combine sensor data with a priori knowledge about the environment the robot travels in.

For example it is often assumed that a detailed map of the area is known. Thus, the problem of identifying the position of the robot is the problem of finding an area within the map such that the expected input from the sensors is at all times in accordance with the actual readings.

Many mobile robot applications depend on the ability of the robot to localize using sparse and many times uncertain information from the environment concerning its position. For example, fusion of odometric and/or inertial sensors with exteroceptive sensors has received significant attention in the past two decades.

The techniques mostly used to process sensor data from different sources are Kalman filtering and Bayesian estimation suitable for position tracking and hypothesis testing respectively.

There have been few examples of the combination of these two methods. In these cases it is assumed that either the motion tracking sensors are perfect or the exteroceptive sensors are capable of  extracting a detailed and precise layout of the surroundings.

A general framework that subsumes both approaches in a single architecture and exploits noisy kinetic information in order to reduce the uncertainty associated with the presence or absence of coarse, reappearing features of the environment is within the focus of the proposed study.
Error state formulation of the underlying Kalman filter based on sensor modeling is mostly appropriate for the localization task.
Restructuring of the exact framework provides for multiple model adaptive estimation.
Full state formulation of the Kalman filter based on simple kinematic modeling is used for the fault detection and identification task.

The intermittent nature of the externally provided or extracted absolute attitude and/or positioning information imposes the system being observable only at certain time instants.
This limitation restricts the potential autonomy of a robot and causes the loss of precious information whenever real-time filtering is involved.
Smoothing of the attitude estimates reduces the overall uncertainty and allows for longer traverses before a new orientation measurement is reqpiuired.
Similar post-processing of the position estimates supports reconstruction of a robot's trajectory and increases the accuracy level during mapping tasks.

Finally, the formulation and treatement of the cooperative multiple robot localization problem is to be examined in this investigation.