1. Repositório Aberto da Universidade do Porto
2. FEUP - Faculdade de Engenharia
3. FEUP - Dissertação

Author(s):	Roberto de Nóbrega Nogueira
Title:	Self-adaptive Cobots in Cyber-Physical Production Systems
Issue Date:	2019-02-07
Description:	The Research Center for Systems and Technologies (SYSTEC) is a research unit hosted by FEUP - Faculdade de Engenharia da Universidade do Porto and by the ISR - Institute of Systems and Robotics. SYSTEC performs fundamental and applied research and development to deepen the knowledge in the areas of systems, control, optimization, estimation, robotics, networked and vehicle systems, power electronics, energy and advanced manufacturing. The Thematic line SYSTEC- MANUFACTURING targets the design, implementation and validation of smart components for advanced manufacturing system that introduce intelligence into industrial processes and contribute to the emergence of the factories of the future, by implementing novel approaches related with Cyber-Physical Production Systems. Targeting added value manufacturing concepts, SYSTEC-MANUFACTURING is highly interdisciplinary as it spans from the design and implementation of the smart components and supporting services (hardware and software), including the required infra-structures, to the CPPS conceptual frameworks underlying control, data analytics, intelligence and coordination. SYSTEC- MANUFACTURING is involved in numerous national and international research and innovation projects and is involved in intensive collaboration with companies - Harms & Wende (DE), Kuka (DE), AWL (NL), TECNAX (FR), PSA (PT), MOTOFIL (PT) - research centres - Centro Ricerche FIAT (IT), Fraunhofer IPA (DE), TECNALIA (SP), ITEA-CNR (IT) - and universities - KTH (SE), TU Delft (NL), University of Karlsruhe (DE), University of Loughborough (UK). Researchers, junior researchers and research interns are integrated in the project teams and have the opportunity to network with fellow researchers from other national and international organisations. Since the introduction of mass production, the level of automation used within manufacturing processes is increasing greatly. Facing the new Industry 4.0 paradigm, industries are shifting towards human-robot collaboration, where the robot, in order to perform a task, is sensitive towards the human operator's actions. Until now, the strategy used in semi-automated processes consists in building static programs to control the robot. These programs have specific instructions that correlate robot sensor data with pre-defined actions. With Cyber-Physical Production Systems (CPPS), the robots should present self- adaptive capabilities in a highly dynamic environment, where robots depend on other robots and human operators, in order to complete a process execution. Human operator's safety is a hot scientific topic in human-robot collaboration environments. One strategy consists in the self-adaptation of the robot in real time, based on the ever changing environment conditions, especially in a shared workspace. The robot should be smart enough to re-plan its movement if an object unpredictably blocks its path or if the human operator's safety is at risk. This dissertation proposal intends to equip a collaborative robotic arm with self-adaptive capabilities, in order to re-plan in real time its task execution, in case of compromised human safety. The student will study artificial vision solutions, which are used to track the human operator's movements when performing tasks and machine learning techniques for workspace modeling and autonomous actuation decision making. The main goal of the present dissertation is to turn a mini robot arm into a functional cobot, by exploring task execution adaptation and movement re-planning, in order to support human safety. The cobot's world perception will be based in artificial vision. It is intended for the robotic arm to learn the human motion when performing a given task, using vision sensors such as the Kinect sensor. It should be able, based on the human's motion in highly dynamic environments, to self-adapt its movements for task execution, without jeopardizing the human safety. The working environment between human and robot is based in an existing cyber-physical system (CPS) architecture. A human-robot collaboration approach, using artificial vision systems, for self-adaptation of cobots in Cyber-Physical Production Systems. There isn't an implementation where these type of capabilities are compliant to be used within a Cyber-Physical Production System architecture. Artificial vision techniques, based on structure from motion, should be explored for workspace 2D/3D modeling, in order identify the position and motion of both human and robotic arm. This will contribute to keep the human operator safe during the execution of shared tasks with unsupervised robot manipulators.
Subject:	Engenharia electrotécnica, electrónica e informática Electrical engineering, Electronic engineering, Information engineering
Scientific areas:	Ciências da engenharia e tecnologias::Engenharia electrotécnica, electrónica e informática Engineering and technology::Electrical engineering, Electronic engineering, Information engineering
DOI:	10.34626/k2ff-dg73
TID identifier:	202397416
URI:	https://hdl.handle.net/10216/119113
Document Type:	Dissertação
Rights:	openAccess
Appears in Collections:	FEUP - Dissertação

Items in DSpace are protected by copyright, with all rights reserved, unless otherwise indicated.