|
|
|
ITER - Remote Handling |
 ITER is the challenging international project for building the fusion experimental reactor, which aims to study the nuclear fusion as an alternative source of energy.The ITER machine operation and support requires remote handling. In particular, to transport large equipment from Tokamak (the heart of ITER) it is required mobile platforms up to 100 tons with similar dimensions to an autobus moving on scenarios with similar dimensions to a football field. This requirement points out several challenges on mobile robotics fields, such as path planning, path following, sensor fusion, signal processing, control interfaces, virtual reality, tests and experimentation. This work will be followed in collaboration with prestigious research laboratories and international companies. More
|
|
Laser range systems to measure live power installations |
Laser range systems are a powerful tool to measure live power installations, since they allow accurate measurements and inspection from a safe distance without disturbing the regular electrical power exploitation. his project is developed on a joint venture among Instituto Superior Técnico (IST) - Institute for Systems and Robotics (ISR) and Labelec SA, a company of the EDP group (electricity operator). More
|
 |
|
Sick Laser Interface for Matlab |
A range sensor is an important tool in mobile robotics, not only for obstacle avoidance, but also for features recognition. However, the first experiments should be made in simple environments with usefull toolboxes as the ones available in Matlab. This page describes an interface for the measurement system Sick Laser model LMS 200. The code is explained and exemplified. More |
|
|
RESCUE |
|
This project will endow a team of outdoors robots with cooperative navigation capabilities, so as to demonstrate the ability of the robots to act individually and cooperatively in search and rescue-like operation scenarios. Search and Rescue (SAR) operations are a challenging application for Robotics since, due to their nature, they naturally foster advances in Artificial Vision, Navigation in outdoors unstructured environments, Distributed Artificial Intelligence and Intelligent Control. More
|
|
|
Particle Filter |
|
The localization problem solved with a probabilistic approach is robust in face of sensor limitations, sensor noise and environment dynamics. Using a probabilistic approach, the localization turns into an estimation problem with the evaluation and propagation of a probability density function. There are several approaches to the localization problem as the Monte-Carlo simulation method, well known as Particle Filter. The main objective of particle filtering is to track a variable of interest, typically with non-Gaussian and potentially multi-modal probability distribution function pdf.
More |
|
|
Navigation in a Maze |
 |
| The project consisted in the develop and implementation of a navigation algorithm. The main goal was the sensor fusion of odometry and a range finder. The algorithm was implemented on a Nomad Super Scout, based on Iinux using the Sick Laser as range-finder. The navigation algorithm was tested in a maze (about 30 m2) built in indoor environments. In addition, there are some videos to ilustrate the experimental results.More
|
Stereo Binocular |
|
|
The human perception is able to create inside his mind a 3D reconstruction of the environment and that is the main reason of this research. The 3D reconstruction is accomplished using the 2D images obtained by the human eyes (the projection on the retina): Stereo Binocular. In the past century, the researchers have started to work with stereograms when the Stereo (Stereopsis) was taking the first steps. Nowadays the Stereo is a powerful tool in the robot navigation to evaluate the distance to the obstacles. Moreover, it is useful to create a 3D representation of the environment. On the other hand, the implementation involves a rough problem: matching. More (only in portuguese)
|
|
Remote Operation through Internet |
The Remote Operation through Internet was the final project of my graduation in the mobile robot field. Implemented in Java language and based on the client/server architecture this project shows the ability to remotely operate a mobile platform through an Internet channel. It includes two Graphical User Interfaces (GUI): one for the server and the other is a set of tools for client operation. The communication requires security and for that reason an encryption algorithm was applied using the common blowfish algorithm.
More (only in portuguese) |
 |
SimProEpi - Simulation of an Epidemic Disease |
|
Based on Biomedial Enginner, the SimProEpi software simulates the evolution of an epidemic disease (caused by a virus). Implemented in Matlab using Automata Cellular Machines, it includes the suitable features of the environment (population attributes, pollution, weather, etc) to the virus propagation. The application runs using a 2D map as well as the main properties of the virus. |
 |
Image Processing |
|
The first experiment using sensor vision was developed under this probject. The main goal is to assemble the control system to the image processing. The real application consists of driving two little carts on a small motor-racing. All the data information comes from a video camera installed in the room. An Interface was implemented on Matlab to follow the main steps of the algorithm (referential transformation, edges detection, car detection, control values, etc) and to evaluate the results. |
|
Control Systems
|
This is an example of different types of control (P, PD, PID and non-linear control) applied on a vehicle to reach the goal. The project includes a vision camera simulation that acquires the robot position. The implementation requires a transformation between two referential sistems.
|
 |
Genetic Algorithms
|
|
As the natural world, the Genetic Algorithms (GA) aims to be a good approximation of a population. That was a pleasant experience applying the GA on maths. Implemented in Matlab the program identifies the roots of the most different types of mathematical functions. The GA includes the main steps of the reproduction: crossover, mutation and selection. That requires a codification for the chromosomes. The selection is accomplished using a fitness function. |
 |
|
|
