The conventional way to develop press-formed metallic components requires a burdensome trial-and-... more The conventional way to develop press-formed metallic components requires a burdensome trial-and-error process for setting-up the technology, whose success depends largely on the operator’s skill and experience. The finite element (FE) simulations of a sheet-metal-forming process help a manufacturing engineer to design a forming process by shifting the costly press-shop try-outs to the computer-aided design environment. Numerical simulations of a manufacturing process, such as rubber-pad forming, have been introduced in order to avoid the trial-and-error procedure and shorten the development phases when tight times-to-market are demanded. The main aim of the investigation presented in this paper was to develop a numerical model that would be able to successfully simulate a rubber-pad forming process. The finite-element method was used for blankand rubber-behavior predictions during the process. The study was concerned with a simulation and investigation of significant parameters (su...
Group technology based manufacturing systems offer the advantages of flow production as well as t... more Group technology based manufacturing systems offer the advantages of flow production as well as the production flexibility of batch manufacturing. In this paper, by employing new clustering techniques, the part-machine spectrum of the manufacturing system and the relevant manufacturing process are analyzed according to design, similarity of machining and product flow. This leads to an organization of the production system into self-contained and self-regulating groups of machines called machine cells. Each machine cell undertakes a maximal production of a family of parts having similar manufacturing characteristics. This paper carried out the ART-1 neural network approach in the analysis of the manufacturing similarity, and modified the basic approach to increase the efficiency of the classification procedure. Developed program packages ART-1 Simulator and FLEXY are used to create part families and machine cells within the group technology design.
Application of ”ART SIMULATOR” for Manufacturing Similarity Identification in Group Technology Design
This chapter carried out the exceptional implementation of ART-1 neural network in the analysis o... more This chapter carried out the exceptional implementation of ART-1 neural network in the analysis of the manufacturing similarity of the cylindrical parts within the group technology design. Established concept of the group technology design begins from the complex part of the group or the group representative. Group representative has all the geometrical elements of the parts in group, and manufacturing procedure may be applied to the machining of any part in the group. The complex part may be realistic or a hypothetical one. The ART-1 artificial neural network provided manufacturing classification according to the geometrical similarities of workpieces for the group of cylindrical parts. For the manufacturing similarity identification within the group technology design, software package "ART Simulator" is developed and presented in this chapter.
To produce a new component there are a trial and error stage to obtain a part without defects, wh... more To produce a new component there are a trial and error stage to obtain a part without defects, which strongly depends on operator's experience. At this stage, the experience of designer and manufacturers should give an important aid to reduce trials to realize the minimization of response time and cost with maximization of the product equality. Rubber pad forming highly improves the formability of the blank because the contact surfaces between the rigid die and the rubber pad is flexible. This method enables to produce sheet metal components with complex contour by relatively low cost because only one rigid die is required. However, in aerospace field, parts are produced in small sheet metal bending and small number of components. It means that it is not worthy for large investments in tools and process design. For these reasons, it is necessary to use FE simulation of manufacturing process during the conceptual design.
It is known that most of the key problems in visual servo control of robots are related to the pe... more It is known that most of the key problems in visual servo control of robots are related to the performance analysis of the system considering measurement and modeling errors. In this paper, the development and performance evaluation of a novel intelligent visual servo controller for a robot manipulator using neural network Reinforcement Learning is presented. By implementing machine learning techniques into the vision based control scheme, the robot is enabled to improve its performance online and to adapt to the changing conditions in the environment. Two different temporal difference algorithms (Q-learning and SARSA) coupled with neural networks are developed and tested through different visual control scenarios. A database of representative learning samples is employed so as to speed up the convergence of the neural network and real-time learning of robot behavior. Moreover, the visual servoing task is divided into two steps in order to ensure the visibility of the features: in the first step centering behavior of the robot is conducted using neural network Reinforcement Learning controller, while the second step involves switching control between the traditional Image Based Visual Servoing and the neural network Reinforcement Learning for enabling approaching behavior of the manipulator. The correction in robot motion is achieved with the definition of the areas of interest for the image features independently in both control steps. Various simulations are developed in order to present the robustness of the developed system regarding calibration error, modeling error, and image noise. In addition, a comparison with the traditional Image Based Visual Servoing is presented. Real world experiments on a robot manipulator with the low cost vision system demonstrate the effectiveness of the proposed approach.
Today's industrial robot systems intensively include external sensors like cameras used for ident... more Today's industrial robot systems intensively include external sensors like cameras used for identification of objects in the working environment of industrial robot. Including cameras in the system of an industrial robot, the control problem of such learning industrial robot is set. Using empirical control strategy based on application of artificial neural networks system, the learning industrial robot can realize adaptive behaviour in the sense of flexible adjustment to changes in the working environment. Unlike natural systems which could learn on the basis of experience, artificial systems are thought to be unable to do so for a long time. However, the concept of empirical control realizes the ability of machine learning on the basis of experience. This paper aims to show that it is possible to realize the empirical control strategy for learning industrial robot using camera and system of artificial neural networks. Results obtained by the system of neural nets have shown that the robot can move the end-effector to the desired location of the object, even in the case where the location differs slightly from the learned patterns.
This paper deals with the integration of Intelligent Mobile Robots (IMR) into the advanced manufa... more This paper deals with the integration of Intelligent Mobile Robots (IMR) into the advanced manufacturing paradigm known as Intelligent Manufacturing Systems (IMS) and analyses some of the most important issues facing this advanced integration. The paper introduces a localization problem of mobile robots and provides a simulation procedure carried out in Matlab ® environment. The presented software application (the source code) provides a chance for modelling a localization problem on the real shop floor. Finally, the first step towards an experimental setup using the LEGO ® Mindstorms TM NXT robotic kit is presented. Although this approach based on simulation may be seen as neglecting essential issues by introducing assumptions common in real problem modelling, it still provides a chance for proper development of the model.
This paper presents methodology for development of software application for integration of proces... more This paper presents methodology for development of software application for integration of process planning, scheduling, and the mobile robot navigation in manufacturing environment. Proposed methodology is based on the Russian Theory of Inventive Problem Solving (TRIZ) and multiagent system (MAS). Contradiction matrix and inventive principles are proved as effective TRIZ tool to solve contradictions during conceptual phase of software development. The proposed MAS architecture consists of six intelligent agents: job agent, machine agent, optimization agent, path planning agent, machine learning agent and mobile robot agent. All agents work together to perform process plans optimization, schedule plans optimization, optimal path that mobile robot follows and classification of objects in a manufacturing environment. Experimental results show that developed software can be used for proposed integration in order to improve performance of intelligent manufacturing systems.
This paper presents the authors' efforts to conceptual design of control system that can learn fr... more This paper presents the authors' efforts to conceptual design of control system that can learn from its own experience. The ability of adaptive behaviour regarding the given task in real, unpredictable conditions is one of the main demands for every intelligent robotic system. To solve this problem, the authors suggest a learning approach that combines empirical control strategy, reinforcement learning and axiomatic design theory. The proposed concept uses best features of mentioned theoretical approaches to produce optimal action in the current state of the mobile robot. In this paper empirical control theory imparts the basis of conceptual solution for the navigation problem of mobile robot. Reinforcement learning enables the mechanisms that memorize and update environment responses, and combining with the empirical control theory determines best possible action according to the present circumstances. Axiomatic design theory accurately defines the problem and possible solution for the given task in terms of the elements defined by two previously mentioned approaches. Part of the proposed algorithm was implemented on the LEGO Mindstorms NXT mobile robot for the navigation task in an unknown manufacturing environment. Experimental results have shown good perspective for development of efficient and adaptable control system, which could lead to autonomous mobile robot behaviour.
Iranian Journal of Science and Technology Transaction B Engineering, Apr 1, 2012
Automated Guided Vehicles (AGV) have been a conventional solution and choice made by many manufac... more Automated Guided Vehicles (AGV) have been a conventional solution and choice made by many manufacturing enterprises as means for Flexible Material Handling Systems (FMHS). In recent years, a considerable number of these vehicles have been installed on shop floors worldwide, effectively proving the usefulness of material handling systems. However, the increasing complexity of demand as well as a need for "make to order" rather than "make to stock" policy implies usage of more intelligent material handling solutions. This paper discusses the usage of an intelligent AGV as means for FMHS which should have Intelligent Material Handling System (IMHS) as the final outcome. This paper presents the experimental results of hybrid robotic control architecture. To evaluate the performance of the architecture, a mobile robot built on LEGO ® Mindstorms NXT technology was used. Some of the architectural modules are based on the implementation of Artificial Neural Networks in order to achieve the needed robustness in exploitation. Then, through simulation using AnyLogic ® 6 software, the performance in terms of manufacturing system effectiveness and workstation utilization is analyzed. Based on the experimental results, the proposed IMHS can have significant advantages over conventional material handling systems.
In this paper we present applied physics research results of gas-permeable contact lenses (CL) th... more In this paper we present applied physics research results of gas-permeable contact lenses (CL) that are manufactured from fluorosilicone acrylate based material (Boston™ type). During contact lenses production the conformation states of polymers belonging to near surface layers of CL surface are changed. Since CL quality crucially depends on surface roughness and optical properties, the properties of surface molecules conformation state and their orientation come into perspective as important factors acting on the molecular level. Therefore, we investigated CL surface by phase contrast atomic force microscopy (PC-AFM), magnetic force microscopy (MFM), and optomagnetic fingerprint (OMF) technique and found out that surface quality and magnetic properties of contact lenses have influence on physical properties of light transmission and that these changes can be detected on the nanolevel of magnetism, as well as optomagnetism. These results carry important 826 D. Stamenković et al.
The International Journal of Advanced Manufacturing Technology, 2013
ABSTRACT Automated guided vehicles (AGVs) are a common choice made by many companies for material... more ABSTRACT Automated guided vehicles (AGVs) are a common choice made by many companies for material handling (MH) in manufacturing systems. AGV-based internal transport of raw materials, goods, and parts is becoming improved with advances in technology. Demands for fast, efficient, and reliable transport imply the usage of the flexible AGVs with onboard sensing and special kinds of algorithms needed for daily operations. So far, the majority of these transport solutions have not considered the modern techniques for visual servoing, monocular SLAM, and consequently, the usage of camera as onboard sensor for AGVs. In this research, a new hybrid control of AGV is proposed. The main control algorithm consists of two independent control loops: position-based control (PBC) for global navigation and image based visual seroving (IBVS) for fine motions needed for accurate steering towards loading/unloading point. By separating the initial transportation task into two parts (global navigation towards the goal pose near the loading/unloading point and fine motion from the goal pose to the loading/unloading point), the proposed hybrid control bypasses the need for artificial landmarks or accurate map of the environment. The state estimation of the robot pose is determined in terms of monocular SLAM, via extended Kalman filter coupled with feedforward neural network—the neural extended Kalman filter (NEKF). NEKF is used to model unknown disturbances and to improve the robot state transition model. The integration of the new hybrid control and NEKF has been tested in laboratory with the mobile robot and simple camera. Experimental results present the effectiveness of the proposed hybrid control approach.
Automated feature recognition (AFR) has provided the greatest contribution to fully automated CAP... more Automated feature recognition (AFR) has provided the greatest contribution to fully automated CAPP system development. The objective of this paper is to review various approaches for solving three major AFR problems: (i) extraction of geometric primitives from a CAD model; (ii) defining a suitable part representation for form feature identification; and (iii) feature pattern matching/recognition. A novel, detailed classification of developed AFR systems has been introduced. This paper also provides a thorough investigation of methods for geometric feature extraction, emphasizing STEP standard application and, finally, a review of recent research reports in the field of AFR with rule-based feature pattern recognition. We discuss potentials and limitations of these approaches and emphasize directions for further research work. #
Uploads
Papers by Bojan Babic