To Örebro University

This paper presents an approach for tracking multiple persons using a combination of colour and thermal vision sensors on a mobile robot. First, an adaptive colour model is incorporated into the measurement model of the tracker. Second, a new approach for detecting occlusions is introduced, using a machine learning classifier for pairwise comparison of persons (classifying which one is in front of the other). Third, explicit occlusion handling is then incorporated into the tracker.

This paper presents an image-based approach for localization in non-static environments using local feature descriptors, and its experimental evaluation in a large, dynamic, populated environment where the time interval between the collected data sets is up to two months. By using local features together with panoramic images, robustness and invariance to large changes in the environment can be handled. Results from global place recognition with no evidence accumulation and a Monte Carlo localization method are shown. To test the approach even further, experiments were conducted with up to 90% virtual occlusion in addition to the dynamic changes in the environment

This paper investigates a learning system based on growing Radial Basis Function (RBF) networks for acquiring reactive behaviours in mobile robotics. The learning algorithm integrates unsupervised and supervised learning, directly mapping the sensor information to the required motor action. The learning system is evaluated through a number of experiments on a real robot. The experimental results show that our learning system can learn a wide range of robot behaviours from simple tasks to complex tasks and demonstrate that the task need not be known at the programming time. This means that many different behaviours could potentially be acquired by the same learning architecture, thus dramatically reducing the development cost of autonomous robotic systems

This paper addresses the problem of estimating proximity to a gas source using concentration measurements. In particular, we consider the problem of gas source declaration by a mobile robot equipped with metal oxide sensors in a turbulent indoor environment. While previous work has shown that machine learning classifiers can be trained to detect close proximity to a gas source, it is difficult to interpret the learned models. This paper investigates possible underlying indicators of gas source proximity, comparing three different statistics derived from the sensor measurements of the robot. A correlation analysis of 1056 trials showed that response variance (measured as standard deviation) was a better indicator than average response. An improved result was obtained when the standard deviation was normalized to the average response for each trial, a strategy that also reduces calibration problems.

This paper presents a learning system that uses Q-learning with a resource allocating network (RAN) for behavior learning in mobile robotics. The RAN is used as a function approximator, and Q-learning is used to learn the control policy in `off-policy' fashion that enables learning to be bootstrapped by a prior knowledge controller, thus speeding up the reinforcement learning. Our approach is verified on a PeopleBot robot executing a visual servoing based docking behavior in which the robot is required to reach a goal pose. Further experiments show that the RAN network can also be used for supervised learning prior to reinforcement learning in a layered architecture, thus further improving the performance of the docking behavior.

In this paper we propose a system consisting of a manipulator equipped with range sensors, that is instructed to follow a trajectory demonstrated by a human teacher wearing a motion capturing device. During the demonstration a three dimensional occupancy grid of the environment is built using the range sensor information and the trajectory. The demonstration is followed by an exploration phase, where the robot undergoes self-improvement of the task, during which the occupancy grid is used to avoid collisions. In parallel a reinforcement learning (RL) agent, biased by the demonstration, learns a point-to-point task policy. When changes occur in the workspace, both the occupancy grid and the learned policy will be updated online by the system.

This paper presents the basis of a new system for making detailed 3D models of underground tunnels. The system is to be used for automated control of mining vehicles. We describe some alternative methods for matching several partial scans, and their applicability for making a complete model of a mine environment

The ICP algorithm and its derivatives is the de facto standard for registration of 3D range-finder scans today. This paper presents a quantitative comparison between ICP and 3D NDT, a novel approach based on the normal distributions transform. The new method ad- dresses two of the main problems of ICP: the fact that it does not make use of the local surface shape and the computationally demanding nearest-neighbour search. The results show that 3D NDT produces accurate results much faster, though it is more sensitive to error in the initial pose estimate.

This paper addresses the problem of simultaneous localisation and mapping (SLAM) by a mobile robot. An incremental SLAM algorithm is introduced that is derived from multigrid methods used for solving partial differential equations. The approach improves on the performance of previous relaxation methods for robot mapping because it optimizes the map at multiple levels of resolution. The resulting algorithm has an update time that is linear in the number of estimated features for typical indoor environments, even when closing very large loops, and offers advantages in handling non-linearities compared to other SLAM algorithms. Experimental comparisons with alternative algorithms using two well-known data sets and mapping results on a real robot are also presented