Tenured Position at the Faculty of Electrical Engineering and Computation of UNICAMP

Assistive Mobile Robot Navigation in a Smart EnvironmentThis project has as its main objective to design, implement, and evaluate a smart environment to help the navigation of assistive robots carrying disabled persons. It has three main lines of research: neurology, digital signal processing, and mobile robotics. The ultimate goal is to build an assistive mobile robot (e.g., a robotized wheelchair) that can navigate with signals acquired from the operator and from the environment. Example of signals generated by the operator include brain activity captured by brain headsets, eye and face tracking acquired from cameras, and whispers acquired from microphones. Once acquired, signals must be processed by a classification system in order to extract high level commands. Such commands are submitted to the robotic vehicle that in its turn translate them into movements. The robot must have high autonomy to navigate safely in crowded areas, avoiding transient and permanent obstacles.My personal focus is to design, implement, and evaluate a software platform from which smart environments that help the navigation of assistive robots carrying disabled persons can be built. The platform will work on both virtual and real environments.Specific Goals:_ To integrate into the software platform the algorithms developed by the digital signal processing team and by the mobile robotics team._ To develop a desktop training system that allows operators to drive simulated robots in different environments such as homes, public places, parks, etc._ To evaluate the platform in real world scenarios with the assistance of researchers from FCM/Unicamp._ To evaluate methodologically patients under neurological treatment._ To prepare reports, presentations, conference papers and articles, and to advise students in topics related to this research.

Design and development of Quince rescue RobotQuince is a small, light-weight 6 degrees of freedom rescue robot designed to address uneven terrain, especially stair steps, step fields and rubbles. Its high degree of mobility is achieved by its mechanical design. It mainly consists of a full waterproof body crawler, four autonomous sub crawlers, a low center of gravity, and powerful brushless DC motors.A modified version of Quince rescue robot is the Japanese robot in use in the crippled Fukushima nuclear plant. It has been modified to sustain radiation, to be teleoperated, to map and to measure radiation levels in the nuclear reactor. Besides being a member of the staff who designed and built Quince modular hardware and software, my focus was to implement a semi-autonomous control of the flippers to help the operator driving Quince. The operator does not need to focus on the control of the flippers when driving, they adapt autonomously to the environment to optimize the traction of the rover. The operator provides only the heading of the robot. This is particularly useful when climbing stairs or over a rough terrain.

A Novel Teleoperated Hybrid Wheel-Limbed Hexapod for the Exploration of Lunar Challenging TerrainsMy research interests concern the field of telerobotic platforms and mobile robots for space exploration. My recent research concerns hybrid leg / wheels robot on unstructured environments and its modular telerobotic platforms. I am focusing on the development of the Lunar Exploration Omnidirectional Netbot (LEON), a novel hybrid wheel / legs hexapod that folds two of its limbs to transform them into wheels. To teleoperate / supervise or simulate LEON or any mobile robot, I am developing ERode Telerobotic dynamic engine based modular platform that integrate a path/action planner.More at: http://www.astro.mech.tohoku.ac.jp/~eric

Study of A High Level Teleoperation Platform for Space Robotic Missions applied to an asteroid surface exploration robot mission (Hayabusa Mk 2): From rescuing teleoperated mission to space exploration one, time delays can drastically extend especially for this target mission where the distance between Earth and the asteroid induce several minutes to get feedback. The robot is to achieve scientific investigation and mapping of the asteroid surface at several locations with fine positioning capability after a large stride movement or a touchdown sampling sequence of the main spacecraft orbiting the minor body. The considered structure of the robot is formed by a central body, with a hexagonal shape and six similar legs, symmetrically distributed around the body. Laser range finders on the body and sensors on the legs to determine force will provide the operator, a model of the local environment of the robot. The proposed platform consists of the following; _ a ground control station, _ an offline simulation environment, _ a working full scale model of the robot hold by a manipulator robot to simulate micro gravity, _the actual robot on the mission.This teleoperation platform is simulation based. The operator is simulating offline the next motion of the robot using the robot replica mounted at the end tip of the PA10 manipulator equipped with a FT sensor, in reconstructed environment. In this way, the hexapod is held and an impedance control is applying the reactions of the environment on the posture, speed and acceleration of the robot as if it was in a micro gravity environment.

Development of networked robotic system intended to be deployed at disaster areas, (SCOPE project) :The system is formed by three mobile robots: two twin crawlers that will have search-and-recognition tasks gathering information about their surroundings, and an outdoor wheeled rover that will approach the area and which will also carry the two crawlers. The communication system for these robots consists of a wireless local area network that will have an operator located at a safe distance controlling them through a satellite-based IP communication, linked to the Japanese satellite ETS-VIII. In order to be able to navigate the crawlers remotely, a wireless LAN camera and a Laser Range Finder (LRF) sensor have been mounted on both of the crawlers. These LRFs scan the area where the crawler is at, obtaining a detailed point-based 3D image.The developed teleoperation platform allows a manual control of multiple robots through a Mixed Reality feedback. The use of both video and a 3D reconstructed environment model allows the operator to navigate in the disaster scene. The video feedback provides a view of dynamical changes of the environment but is not primordial to the navigation. A drawback of a current platform, however, is that the direct navigation through it becomes impossible as the delay in the media increases, preventing its use in case of space missions where the range of time delay between earth and the mission is minutes or hours.