1,720,964 research outputs found
Nuove Tecniche di Stima e Controllo per Robot Mobili nel Mondo Reale
No full textQuesta tesi tratta del trasferimento tecnologico delle più recenti scoperte nel settore della robotica mobile in applicazioni industriali e di servizio nel mondo reale. Nello specifico, le ricerche presentate sono state svolte nel framework di due progetti: il progetto DiRò e il progetto REMOCAL.
Lo scopo del progetto DiRò (Distretto della Robotica Mobile) è di creare una collaborazione fra alcune aziende della provincia di Reggio Emilia con l’obiettivo di sviluppare nuove soluzioni nell’ambito della robotica mobile. Questa tesi tratta principalmente di tre obiettivi realizzativi di questo progetto.
In questa tesi verrà per prima cosa affrontata la progettazione di un tosaerba automatico e, in particolare, del suo algoritmo di localizzazione. Il robot usa una particolare implementazione dell’ Extended Kalman Filter (EKF) per localizzarsi nel giardino. Quindi è implementata una innovativa strategia di navigazione che fa sì che il robot possa affrontare il suo compito di tosatura assicurando una copertura uniforme e completa del prato.
Dopodiché viene discussa la caratterizzazione di sensori di visione 3D. In particolare è presentato il confronto di una stereo–camera (Bumblebee2) e un sensore RGB–D (il Kinect Microsoft). Le performance dei due sensori sono testate con un algoritmo di localizzazione basato su un EKF.
La discussione sul progetto DiRò termina con la progettazione di un innovativo sistema per la stima dello Stato di Carica di batterie al piombo, anch’esso basato sull’ EKF.
Infine viene descritto il progetto REMOCAL (Robotic Equipment for Measuring by Optics Calibrator). Esso consiste nella progettazione e nella realizzazione di un robot mobile capace di calibrare autonomamente un dispositivo di misura della convergenza dei pneumatici per automobili. In questo documento sono riportati sia la progettazione meccanica che la progettazione del software di controllo del robot.This thesis deals with the technology transfer of the most recent findings in the mobile robotics field into applications into real world industries and service applications. Specifically the presented research has been done in the framework of two projects: DiRò and REMOCAL.
The main scope of the DiRò (Distretto della Robotica Mobile) project is to create a collaboration between companies of the province of Reggio Emilia with the objective to develop new solution in the framework of the Mobile Robotics. This thesis focuses mainly on three development objectives of this project.
First the design of a smart autonomous lawnmower, and in particular its localization algorithm, is described in this thesis. The robot uses a particular implementation of the Extended Kalman Filter (EKF) for localizing itself in the garden. Then a novel navigation strategy has been implemented, in order to make the robot doing its mowing task ensuring a complete and even coverage of the assigned working area.
Then the characterization of 3D sensors is discussed. In particular a comparison between a stereo–camera (a Bumblebee2 stereo–camera) and an RGB–D sensor (the Microsoft Kinect) is presented. Then the performance of the two sensors is tested through a localization algorithm based on an EKF.
The Dir`o project discussion eventually ends with the design of a novel lead–acid batteries State of Charge (SoC) estimator also based on the EKF algorithm.
Finally the REMOCAL (Robotic Equipment for Measuring by Optics Calibrator) project is described. The project consists of the design and the realization of a mobile robot able to autonomously calibrate a car tyres convergence measurement device. In this manuscript both the mechanical design of the robot and the development of the software which controls the robot are reported
A Novel Inertial/RFID Based Localization Method for Autonomous Lawnmowers
No full textRobotic lawnmowers currently available in the market cover their assigned area using a random reflection navigation strategy. While this strategy has been widely accepted for autonomous vacuum cleaning systems, it poses quality problems in outdoor applications since a randomic crossing of the garden can lead to an uneven mowing. In this paper we propose a localization algorithm based on a modified Constrained Kalman Filter that allows to implement an efficient navigation strategy and to increase the quality of service of the mower. This method properly merges data coming from an Inertial Measurement Unit (IMU) and from an RFID (Radio-Frequency IDentification) antenna with information given by the Hall sensors of the wheels of the robot. The proposed algorithm has been verified first by simulation, and then with experiments by building a prototype lawnmower robot
Implementation of Coordinated Complex Dynamic Behaviors in Multirobot Systems
No full textDecentralized control strategies for multirobot systems have been extensively studied over the past few years. Typically, these strategies aim at exploiting local interaction rules to regulate the overall state of the multirobot system toward a desired configuration, thus generating some desired coordinated behaviors, such as synchronization, swarming, deployment, or formation control. However, when considering the real-world application of multirobot systems, more complex cooperative dynamic behaviors are desirable. Along these lines, in this paper, we propose a methodology to control a multirobot system for cooperatively tracking arbitrarily defined periodic setpoint trajectories. This objective is fulfilled partitioning the multirobot system into independent robots (that can provide control inputs) and dependent robots (that are controlled through local interaction). The motion of the independent robots is then defined in such a way that, exploiting local interactions, the dependent robots are controlled to track the desired trajectories. The proposed control strategy is validated by means of simulations and experiments on real robots
Decentralized Control of Cooperative Robotic Systems for Arbitrary Setpoint Tracking while Avoiding Collisions
No full textDecentralized control strategies for multi-robot systems have been extensively studied in the literature, with the aim of regulating the overall state of the system to some desired configuration. Recently a few works have appeared that propose methodologies for solving more complex problems: in particular, partitioning of the multi-robot system into two subgroups can be exploited for cooperatively tracking arbitrarily defined setpoint functions. Specifically, a small number of independent robots can be exploited for controlling the overall state of the dependent robots. Based on this formulation, in this paper we exploit the null space based behavioral approach to obtain setpoint tracking for a multi-robot system, while avoiding inter-robot collisions
A low cost localization algorithm for an autonomous lawnmower
No full textThis paper aims at implementing a localization algorithm, based on a Kalman filter, in order to localize an autonomous lawnmower which operates in an outdoor environment using only low cost sensors. In particular the position and the orientation of the robot are estimated using an IMU (Inertial Measurement Unit), an RFID (Radio Frequency IDentification) antenna and an RFID reader (which reads the presence of RFID tags scattered on the border of the lawn to be mowed) and a Wifi antenna, on-board the robot, which measures the RSSI (Received Signal Strength Indicator) of the signal sent by other Wifi end devices positioned in the working area in known positions. The efficiency of the proposed algorithm is then tested first through simulations and then experimentally on a prototype lawnmower
Coordinated dynamic behaviors in multi-robot systems with time-varying topologies
No full textA control strategy for the execution of coordinated complex dynamic behaviors in multi-robot systems is introduced in this paper. In particular, considering a dependent-independent robot partitioning of a multi-robot system, we introduce a methodology for controlling the independent robots in such a way that, exploiting local interaction, the dependent robots are driven to track desired periodic setpoint trajectories. The control strategy is designed taking explicitly into account variations in the interconnection topology, due to relative movements among the robots
TIREBOT: A novel tire workshop assistant robot
No full textIn this paper a novel tire workshop assistant robot is presented. The robot can load heavy wheels and transport them in any place of the workshop, lifting the operator from the effort of doing it manually. The operator can interact with the robot through gestures, which are captured by a RGB-D camera and translated into commands, or teleoperate it through a control station equipped with a haptic interface, which can also return a force feedback. The robot can also operate autonomously with a mission manager that decides where the robot has to go in order to accomplish a certain task. The paper presents a description of the design process and motivations of every design choice
TIREBOT: A collaborative robot for the tire workshop
No full textCollaborative robots allow to relieve humans from tedious and/or tiring tasks while leaving to the operator high-level activities, where their expertise is needed. In this paper the mechanical design and the collaborative control of TIREBOT, a robotic assistant helping tire workshop operators in the wheel replacement process, is illustrated. The effectiveness of TIREBOT is validated in a real tire-workshop. Furthermore, the possibility of using TIREBOT as an autonomous personal forklift has been validated in an industrial scenario
Safe navigation and experimental evaluation of a novel tire workshop assistant robot
No full textThis paper presents TIREBOT, a novel tire-workshop robotic co-worker that can safely move in a tire workshop and assist the operator in lifting and transporting wheels among several working stations. A safe and cooperative navigation strategy based on the concept of danger field is illustrated. Finally, TIREBOT is experimentally evaluated in a real tire-workshop and used by real operator in order to assess the usability and the effectiveness of the robot in a real operating scenario
Interacting With a Mobile Robot with a Natural Infrastructure-Less Interface
No full textIn this paper we introduce a novel approach that enables users to interact with a mobile robot in a natural manner. The proposed interaction system does not require any specific infrastructure or device, but relies on commonly utilized objects while leaving the user’s hands free. Specifically, we propose to utilize a smartwatch (or a sensorized wristband) for recognizing the motion of the user’s forearm. Measurements of accelerations and angular velocities are exploited to recognize user’s gestures and define velocity commands for the robot. The proposed interaction system is evaluated experimentally with different users controlling a mobile robot and compared to the use of a remote control device for the teleoperation of robots. Results show that the usability and effectiveness of the proposed natural interaction system based on the use of a smartwatch provide significant improvement in the human-robot interaction experience
- …
