Lucas Molina

Fillet Weld Identification for Automatic Inspection of Spherical Tanks

In this article the development of a system for fillet weld identification is considered for automatic inspection of storage spheres. Distance sensors based on different physical principles are characterized and selected, thus getting, greater confidence of the acquired data. The data provided from the different types of sensors will be submitted to a fusing process, aiming at to increase the confidence and performance of the system.

Fault-tolerant weld line detection for automatic inspection of storage tanks based on distance and visual information fusion

Quality control, cost reduction and above all, human and environmental safety are great reasons that stimulate the investments in technologies like automatic inspection. The automatic inspection of weld lines in storage tanks is of special interest, due to the fact that such tanks are currently used to store harmful products. For a reliable inspection it is necessary to accurate detect the weld line position. In this paper the development of a system to perform weld line detection in storage tanks is proposed. Two redundant systems, based on different physical principles, distance and visual information, are implemented and tested. Such systems make use of a fault-tolerant estimation process based on the α-β filter. Finally, the outputs of the two redundant systems are fused, aiming at to increase the confidence and performance of the system.

A Model-Based Fault-Tolerant Weld Line Detection for Automatic Inspection of Storage Tanks Using Visual Information and a-ß Filter

Quality control, cost reduction and above all, human and environmental safety are great reasons that stimulate the investments in technologies such as automatic inspection. The automatic inspection of weld lines in storage tanks is of special interest due to the fact that such tanks are currently used to store harmful products. For a reliable inspection it is necessary to accurately detect the weld line position. In this paper the development of a system to perform weld line detection in storage tanks is proposed. The system based on visual information is implemented and tested. Such system makes use of a fault-tolerant estimation process based on the α-β filter and a model-based fault detection method, proposed in this paper.

A Methodology to Designing Strategies for Robot Soccer Based on Discrete Event Systems Formalism

In the robot soccer, a good strategy is essential to the success of a team. It is responsible to define the behavior of each robot on the many game situations and make the team play in a cooperative manner to win the game. To accomplish this, it is proposed in this paper a methodology to designing strategies for robot soccer based on discrete event systems formalism. The main idea is to allow the construction of a game strategy where the behavior of each robot can be dynamically modified, depending on the context of the match. This strategy is modeled based on the discrete events system (DES) theory. The DES formalism enables to implement the interaction between the various behaviors of a system in a simple way, allowing the addition of new behavior without changing the current behaviors already in use, which is the main advantage of the proposed approach.

Compact RRT: A New Approach for Guided Sampling Applied to Environment Representation and Path Planning in Mobile Robotics

In this paper it is presented a new approach to increase the dispersion of the nodes in the RRT, this approach allows a compact representation of the environment by reducing the nodes redundancy, with this, the number of samples discarded, the computational cost and the processing time of tree growth is also reduced. The developed method performs the polarization of the nodes creating search regions with the highest probability of connectivity to the tree, these regions are created from environmental discretization. Furthermore, the polarization technique presented in this paper is robust to environmental variation, not reducing its performance in more complex environments with narrow passages or long corridors. When using the Compact RRT, path planning stage is faster, because the number of nodes in the tree will be lower compared to Classic RRT.

Evolutionary robotics applied to the multi-robot worst-case pursuit-evasion problem

This paper proposes the use of an evolutionary robotics approach to solve the worst-case pursuit-evasion problem, in which evaders are considered arbitrarily fast and omniscient, while pursuers have limited sensing and communication capabilities, with no previous knowledge of the environment. Unlike most work in evolutionary robotics, we offer a control system for multiple mobile robots based on finite state machines derived using a genetic algorithm. Results show that, given a sufficient number of robots, the evolved system is capable of clearing a discrete map, including multiply connected maps, of all previous contamination.

RRT-Edge: A compact approach for path planning of mobile robots

In this article, a new method is presented in order to increase node dispersion on RRTs. This approach allows for a better environment representation, reducing the amount of nodes in the tree. The proposed RRT differs from Classic RRT by an insertion of a new node-tree connection method. The proposed affixing rule uses a tree whose edges are of variable length, which more rapidly explores the environments structure. In order to reduce edges overlapping and the amount of nodes, new nodes could be added along the edges, instead of only outside of them. Furthermore, this work proposes a discrete density metric which will be used to compare, by simulation experiments, the classic method to the technique here presented. This technique generates improvements on the uniformity of the trees expansion and on the graphs exploration speed in the path-planning phase. The results obtained suggest a reduction of up to 70% of the node amount, while keeping a constant density.

Heuristics for the Multi-Robot Worst-Case Pursuit-Evasion Problem

There is a growing demand for the use of robots to assist humans in their tasks, especially those involving risks, such as search and rescue. For this reason, coordination among several robots has been a common option, and one of the ways to study and model these applications involves the problem of pursuit evasion. This paper extends the results presented earlier on the use of an evolutionary robotics approach to solve the worst case pursuit-evasion problem, in which evaders are considered arbitrarily fast and omniscient, while pursuers have limited sensing and communication capabilities, with no prior knowledge regarding environments, which are treated as discrete and can be multiply connected. First, a formulation based on random walk is offered. Then, the concept is extended to include a decentralized multi-robot control system based on a finite-state machine with state-action mapping defined by means of a genetic algorithm. Results show that the proposed system is able to decontaminate several types of maps, but does not generalize to all initial conditions, due to the incompleteness in the automaton mapping. Therefore, a complementary approach is presented in which random walk is used alternatively with the evolved automaton, indicating random actions in cases of states not sufficiently visited during evolution. In addition, a comparative analysis of the evolutionary approach and the random walk formulation is also carried out.

A Simple Visual Odometry Approach Based on Point Clouds

In this paper, a simple approach for estimating the ego-motion of a vehicle is proposed. The system can estimate the motion based on a sequence of images and point cloud data. The system is capable of estimate the ego-motion through a 2-D to 3-D mapping of the detected features on the image considering that the calibration parameters of the sensors are available. Singular value decomposition is used to estimate the rigid-body transformation between two point clouds, and we employ a RANSAC outlier rejection method to estimate a more accurate ego-motion. Datasets from the KITTI Benchmark Suite are used in order to evaluate the proposed method. The experiments show that our approach performed well with translational errors below 5%.

Applications of Stochastic Approaches in Mobile Robotics: An Overview of Recent Brazilian Works

Stochastic approaches have become increasingly common in robotics. Techniques that consider the uncertainties of sensors and models are being adopted in various applications. In this context, state estimation techniques play an important role, since they deal directly with these uncertainties, making systems more reliable. This paper aims to survey the state of the art of application of stochastic approaches in robotics in the context of the work developed in Brazil. At the end, the main applications that use stochastic approaches in robotics are summarized, and the main tools used in each case are suggested. Moreover, the main outstanding issues, which should be the focus of further research, are indicated. Thus, this paper is intended to be a useful tool for new researchers in this booming research area.