The ability to simultaneously localize a robot and accurately map its surroundings is considered by many to be a key prerequisite of truly autonomous robots. However, few approaches to this problem scale up to handle the very large number of landmarks present in real environments. Kalman filter-based algorithms, for example, require time quadratic in the number of landmarks to incorporate each sensor observation. This paper presents FastSLAM, an algorithm that recursively estimates the full posterior distribution over robot pose and landmark locations, yet scales logarithmically with the number of landmarks in the map. This algorithm is based on a factorization of the posterior into a product of conditional landmark distributions and a distribution over robot paths. The algorithm has been run successfully on as many as 50,000 landmarks, environments far beyond the reach of previous approaches. Experimental results demonstrate the advantages and limitations of the FastSLAM algorithm on both simulated and real-world data.

This article provides a comprehensive introduction into the field of robotic mapping, with a focus on indoor mapping. It describes and compares various probabilistic techniques, as they are presently being applied to a vast array of mobile robot mapping problems. The history of robotic mapping is also described, along with an extensive list of open research problems.

The Extended Kalman Filter (EKF) has been the de facto approach to the Simultaneous Localization and Mapping (SLAM) problem for nearly fifteen years. However, the EKF has two serious deficiencies that prevent it from being applied to large, realword environments: quadratic complexity and sensitivity to failures in data association. FastSLAM, an alternative approach based on the Rao-Blackwellized Particle Filter, has been shown to scale logarithmically with the number of landmarks in the map [10]. This efficiency enables FastSLAM to be applied to environments far larger than could be handled by the EKF. In this paper, we will show that FastSLAM also substantially outperforms the EKF in environments with ambiguous data association. The performance of the two algorithms is compared on a real-world data set with various levels of odometric noise. In addition, we will show how negative information can be incorporated into FastSLAM in order to improve the accuracy of the estimated map.

We present a comparison of an extended Kalman filter and an adaptation of bundle adjustment from computer vision for mobile robot localization and mapping using a bearing-only sensor. We show results on synthetic and real examples and discuss some advantages and disadvantages of the techniques. The comparison leads to a novel combination of the two techniques which results in computational complexity near Kalman filters and performance near bundle adjustment on the examples shown.

Abstract — We describe a probabilistic framework for detection and modeling of doors from sensor data acquired in corridor environments with mobile robots. The framework captures shape, color, and motion properties of door and wall objects. The probabilistic model is optimized with a version of the expectation maximization algorithm, which segments the environment into door and wall objects and learns their properties. The framework allows the robot to generalize the properties of detected object instances to new object instances. We demonstrate the algorithm on real-world data acquired by a Pioneer robot equipped with a laser range finder and an omni-directional camera. Our results show that our algorithm reliably segments the environment into walls and doors, finding both doors that move and doors that do not move. We show that our approach achieves better results than models that only capture behavior, or only capture appearance. I.

In recent years, particle filters have solved several hard perceptual problems in robotics. Early successes of particle filters were limited to low-dimensional estimation problems, such as the problem of robot localization in environments with known maps. More recently, researchers have begun exploiting structural properties of robotic domains that have led to successful particle filter applications in spaces with as many as 100,000 dimensions. The fact that every model---no mater how detailed---fails to capture the full complexity of even the most simple robotic environments has lead to specific tricks and techniques essential for the success of particle filters in robotic domains. This article surveys some of these recent innovations, and provides pointers to in-depth articles on the use of particle filters in robotics.

This thesis deals with the Simultaneous Localisation and Mapping algorithm as it pertains to the deployment of mobile systems in unknown environments. Simultaneous Localisation and Mapping (SLAM) as defined in this thesis is the process of concurrently building up a map of the environment and using this map to obtain improved estimates of the location of the vehicle. In essence, the vehicle relies on its ability to extract useful navigation information from the data returned by its sensors. The vehicle typically starts at an unknown location with no a priori knowledge of landmark locations. From relative observations of landmarks, it simultaneously computes an estimate of vehicle location and an estimate of landmark locations. While continuing in motion, the vehicle builds a complete map of landmarks and uses these to provide continuous estimates of the vehicle location. The potential for this type of navigation system for autonomous systems operating in unknown environments is enormous. One significant obstacle on the road to the implementation and deployment of large scale SLAM algorithms is the computational effort required to maintain the correlation information between features in the map and between the features and the vehicle. Performing the update of the covariance matrix is of O(n3) for a straightforward implementation of the Kalman Filter. In the case of the SLAM algorithm, this complexity can be reduced to O(n2) given the sparse nature of typical observations. Even so, this implies that the computational effort will grow with the square of the number of features maintained in the map. For maps containing more than a few tens of features, this computational burden will quickly make the update intractable - especially if the observation rates are high. An effective map-management technique is therefore required in order to help manage this complexity. The major contributions of this thesis arise from the formulation of a new approach to the mapping of terrain features that provides improved computational efficiency in the SLAM algorithm. Rather than incorporating every observation directly into the global map of the environment, the Constrained Local Submap Filter (CLSF) relies on creating an independent, local submap of the features in the immediate vicinity of the vehicle. This local submap is then periodically fused into the global map of the environment. This representation is shown to reduce the computational complexity of maintaining the global map estimates as well as improving the data association process by allowing the association decisions to be deferred until an improved local picture of the environment is available. This approach also lends itself well to three natural extensions to the representation that are also outlined in the thesis. These include the prospect of deploying multi-vehicle SLAM, the Constrained Relative Submap Filter and a novel feature initialisation technique. Results of this work are presented both in simulation and using real data collected during deployment of a submersible vehicle equipped with scanning sonar.

This paper presents methods of localization using cooperating landmarks (beacons) that provide the ability to measure range only. Recent advances in radio frequency technology make it possible to measure range between inexpensive beacons and a transponder. Such a method has tremendous benefit since line of sight is not required between the beacons and the transponder, and because the data association problem can be completely avoided. If the positions of the beacons are known, measurements from multiple beacons can be combined using probability grids to provide an accurate estimate of robot location. This estimate can be improved by using Monte Carlo techniques and Kalman filters to incorporate odometry data. Similar methods can be used to solve the simultaneous localization and mapping problem (SLAM) when beacon locations are uncertain. Experimental results are presented for robot localization. Tracking and SLAM algorithms are demonstrated in simulation. 1

In this paper we present a technique for mapping partially observable features from multiple uncertain vantage points. The problem of concurrent mapping and localization (CML) is stated as follows. Starting from an initial known position, a mobile robot travels through a sequence of positions, obtaining a set of sensor measurements at each position. The goal is to process the sensor data to produce an estimate of the trajectory of the robot while concurrently building a map of the environment. In this paper, we describe a generalized framework for CML that incorporates temporal as well as spatial correlations. The representation is expanded to incorporate past vehicle positions in the state vector. Estimates of the correlations between current and previous vehicle states are explicitly maintained. This enables the consistent initialization of map features using data from multiple time steps. Updates to the map and the vehicle trajectory can also be performed in batches of data acquired from multiple vantage points. The method is illustrated with sonar data from a testing tank and via experiments with a B21 land mobile robot, demonstrating the ability to perform CML with sparse and ambiguous data. KEY WORDS—mapping, navigation, mobile robots 1.

This paper presents a novel approach to the Simultaneous Localisation and Mapping (SLAM) algorithm that exploits the manner in which observations are fused into the global map of the environment to manage the computational complexity of the algorithm and improve the data association process. Rather than incorporating every observation directly into the global map of the environment, the Constrained Local Submap Filter (CLSF) relies on creating an independent, local submap of the features in the immediate vicinity of the vehicle. This local submap is then periodically fused into the global map of the environment using appropriately formulated constraints between the common feature estimates. This approach is shown to be effective in reducing the computational complexity of maintaining the global map estimates as well as improving the data association process.