|09:00-09:30 Pozvano predavanje
|Tadej Petrič (Jozef Stefan Institute, Ljubljana, Slovenia)
Robots That Learn Like Humans and from Humans: Towards Efficient Human-Robot Collaboration
|A. Jurić, F. Kendeš, I. Marković, I. Petrović (Faculty of Electrical Engineering and Computing, Zagreb, Croatia)
A Comparison of Graph Optimization Approaches for Pose Estimation in SLAM
Simultaneous localization and mapping (SLAM) is an important tool that enables autonomous navigation of mobile robots through unknown environments. As the name SLAM suggests, it is important to obtain a correct representation of the environment and estimate a correct trajectory of the robot poses in the map. Dominant state-of-the-art approaches solve the pose estimation problem using graph optimization techniques based on the least squares minimization method. Among the most popular approaches are libraries such as g2o, Ceres, GTSAM and SE-Sync. The aim of this paper is to describe these approaches in a unified manner and to evaluate them on an array of publicly available synthetic and real-world pose graph datasets. In the evaluation experiments, the computation time and the value of the objective function of the four optimization libraries are analyzed.
|I. Hrabar, J. Goričanec, Z. Kovačić (University of Zagreb Faculty of Electrical Engineering and Computing, Zagreb, Croatia)
Towards Autonomous Navigation of a Mobile Manipulator in a Steep Slope Vineyard
This paper explores potential applications of an autonomous all-terrain mobile manipulator (AMM) in the agricultural vineyard environment. A robotic arm mounted on a mobile base performs a spraying task while moving past vines and a bud rubbing task while stopping next to vines. We discuss the challenges of AMM navigation when performing tasks in steep vineyard terrain. The presented navigation approach is based on the use of a heterogeneous robotic system consisting of an unmanned aerial vehicle (UAV) and AMM. The navigation information provided by the UAV gives the waypoints for the AMM. We also present preliminary results from navigation simulation, based on digital elevation model (DEM). The model uses real-world data, acquired by UAV photogrammetric measurement campaign. We conclude by highlighting some open problems and related future research directions.
|M. Rezo, K. Čagalj (University of Zagreb Faculty of Electrical Engineering and Comp, Zagreb, Croatia), I. Ušljebrka (Cromaris d.d., Zadar, Croatia), Z. Kovačić (University of Zagreb Faculty of Electrical Engineering and Comp, Zagreb, Croatia)
Collecting Information for Biomass Estimation in Mariculture with a Heterogeneous Robotic System
In this paper, we address the problem of fish stock estimation in marine fisheries using a heterogeneous robotic system consisting of unmanned aerial vehicles (UAVs) and unmanned underwater vehicles (UUVs). UAVs take aerial photographs of the cage during fish feeding, while UUVs take photographs of fish from top to bottom in the cage. The photos and videos obtained provide the input data for estimating the number of fish and the amount of biomass of fish in the cage. The paper analyzes a number of factors that affect the accuracy of the estimate. Preliminary results obtained with an approximate method for estimating the number of fish, based on the processing of images obtained in a virtual simulator and resembling aerial photographs of fish taken during feeding, are described. The results obtained show that this problem is extremely complex and that it is worth trying to use machine learning and artificial intelligence methods to achieve the desired maximum estimation error of less than 20%.
|K. Tešić, A. Baričević, M. Serdar, I. Banjad Pečur (University of Zagreb Faculty of Civil Engineering, Zagreb, Croatia)
Ground Robotic Systems for the Inspection of Concrete Bridges
This paper focuses on reviewing relevant work on autonomous ground systems for concrete bridge inspection. The current inspection of bridges is still based on visual inspection by inspectors or by using semi-destructive techniques. Current inspection practices require a large amount of time for inspection. In addition, complex scaffolding or expensive equipment is required for inaccessible areas, which also poses a risk to the safety of inspectors. These drawbacks could be overcome by using robotic systems equipped with non-destructive techniques (NDT). This paper presents the robotic systems that have been used in the inspection of concrete bridges, mainly for the localization of reinforcement, corrosion assessment and crack detection.
|M. Božić, B. Jerbić, M. Švaco (Faculty of Mechanical Engineering and Naval Architecture, University of Zagreb, Zagreb, Croatia)
Development of a Mobile Wall-Climbing Robot with a Hybrid Adhesion System
In this paper, the development process of a mobile wall-climbing robot (WCR) for inspection tasks of vertical concrete structures is presented. The WCR uses a novel hybrid adhesion system based on a combination of negative pressure and thrust. Negative pressure and horizontal thrust are generated using Electric Ducted Fans (EDF) with a custom shroud design, while vertical thrust is generated by a drone-type propulsion unit. The EDF generates a horizontal force that allows the robot to be in contact with the vertical surface, while the drone-type propulsion unit generates a vertical force that helps the robot overcome the gravitational force. This hybrid adhesion system allows the mobile WCR to adhere to most vertical structures, including concrete columns and infrastructure that the robot will inspect. First, a testing prototype of the mobile WCR with a novel hybrid adhesion system is presented. Then, two experimental setups are presented. In the first setup, the EDF and shrouds were tested in different configurations to maximize the adhesion forces, while in the second setup, the BLDC motors were tested with different types and sizes of propellers to optimize the energy consumption in relation to the generated thrust. Based on the obtained results from these two experimental setups, the final hybrid adhesion system configuration is chosen.
|E. Szymanska (Wroclaw University of Science and Technology, Faculty of Electronics, Wroclaw, Poland), L. Petrovic, I. Markovic, I. Petrovic (University of Zagreb, Faculty of Electrical Engineering and Computing, Zagreb, Croatia)
Mobile Robot Teleoperation via Android Mobile Device with UDP Communication
Teleoperation is an essential component of a robotic system utilized for execution of various remote tasks, e.g. telepresence, reconnaissance or search and rescue. In this paper we investigate the usage of User Datagram Protocol (UDP) for two-way communication between an Android mobile device and Robot Operating System (ROS). UDP offers simplified and connectionless communication, which allows high transmission rates and enables real-time applications. We develop an Android application which turns omnipresent Android devices into cheap and intuitive robot teleoperation interfaces. It enables interpreting user’s desired movement choices into corresponding datagrams, which are then sent to the robot via wireless network. As a case study, we implement a solution for controlling the Turtlebot3 differential drive mobile robot within ROS. We examine the utility of UDP-based teleoperation in both simulation and real-life environment. We complement the paper with modular open-source implementations of our Android application and a ROS package featuring the Turtlebot3 controller.
|J. Grönman, M. Saarivirta (Tampere University, Pori, Finland), T. Aaltonen, T. Kerminen (Satakunta University of Applied Scieces, Pori, Finland)
Review of Artificial Intelligence Applications in the ROS Ecosystem
Artificial intelligence has developed rapidly in recent years. It is permanently established in the field of robotics and science. It is seen as an intelligent decision maker which interacts with the environment or material and can work to achieve the best possible outcome. The biggest factors are the constant increase in computing power and the greater availability of data. With a large amount of data and computational power, solutions to complex problems can be created through machine learning. An application that can independently exploit the information provided to it and make decisions based on it can be considered an AI application. This can be utilized in robotics, for example, by teaching a chess robot to recognize different game pieces from each other. Machine learning can be used to teach a robot to adapt to different situations in a job, for example, to identify chess pieces from different perspectives or to find the most effective solution to a given problem. This article examines the use of artificial intelligence in ROS robotics applications.
|E. Salvato, G. Fenu, E. Medvet, F. Pellegrino (Department of Engineering and architecture, University of Trieste, Trieste, Italy)
Characterization of Modeling Errors Affecting Performances of a Robotics Deep Reinforcement Learning Controller in a Sim-to-Real Transfer
Simulation is a powerful tool used to train Reinforcement Learning (RL) agents involved in robotic tasks.
It allows to collect large amount of data in comparatively faster and safer way than on the real robot.
However, a simulator is only an approximation of the physical system to be controlled.
Due to modeling errors, a controller learned on the simulator dynamics may behave differently once applied to the real robot.
In the worst case, the controller, although being successful when applied on the simulator, may fail when applied on the real platform.
In this paper, we deal with the sim-to-real transfer of a RL controller for a Poppy Ergo Jr robotic arm for a positioning task: i.e., moving the servo joints in order to reach a desired target position with the end-effector.
In particular, we want to investigate the differences between the real robot and its simulator and how they affect the controller performance after its transfer from the simulator to the real platform.
|I. Astrov, A. Udal, M. Jaanus (Tallinn University of Technology, Tallinn, Estonia)
A Model-Based Adaptive Control of an Autonomous Driving Car for Lane Change Maneuver
Design of effective computer control algorithms for different autonomous vehicles has become one of the high priority tasks in the present era of fast development of Cyber-Physical Systems (CPS), Industry 4.0 Systems and the segment of self-driving cars in world economy in general. Paper discusses control mathematics of an Autonomous Vehicle (AV) highly nonlinear model of small shuttle bus weight class. To ensure safety of the autonomous line switching and compliance with the traﬃc rules in demanding urban conditions of a smart city, we offer a specially designed nonlinear controller capable to assure smooth trajectory changes under given spatial constraints. Paper describes design and simulation of controller algorithm in Simulink/MATLAB environment. Results of the design will be implemented and tested on the full-scale autonomous shuttle minibus.
|D. Kudek, M. Sužnjević (Faculty of Electrical Engineering and Computing, Zagreb, Croatia)
Automated Robot Control for a Game of Chess in Unity Game Engine through Artificial Intelligence
The topic of this paper is to study the possibility of using Unity game development engine for robot control. The aim of the work is to create a virtual environment in which the game of chess is simulated, through a duel of two robots controlled by artificial intelligence. As part of the work, real robot models were implemented in the Unity game engine. The simulated robots were ABB’s IRB-120 arms with two joints. The movement of the robot is fully simulated within the physics simulation in the Unity system. The Forward and Backward Reaching Inverse Kinematics (FABRIK) algorithm was used for the inverse kinematics algorithm. For calculating the next move, external artificial intelligence library Stockfish was used and integrated with the Unity game engine. The final application has automated moves between the robots, has the option of a simple change of the viewpoint through camera movement, and is intended to be used in future work for the control of a real robot.
|Z. Gršković, J. Peršić, I. Marković, I. Petrović (Faculty of Electrical Engineering and Computing, Zagreb, Croatia)
Depth from Mono Accuracy Analysis by Changing Camera Parameters in the CARLA Simulator
Depth estimation is an important task in robotics and autonomous driving. By estimating depth and relying only on a single camera, it is no longer necessary to add and calibrate additional sensors – usually a second camera. However, such an approach requires training on extensive datasets and obtaining real-world datasets is time consuming and costly. Given that, using photorealistic simulators can be beneficial, since a multitude of various scenes can be created. In this paper we present an approach to training a deep neural network based on the ResNet architecture for estimating depth from a single camera. We target road vehicle scenes and use the CARLA simulator. We evaluate the trained network on the real-world KITTI dataset images and in the CARLA simulator. In the simulated experiments, we compare the performance with respect to the changes in camera intrinsic and extrinsic calibration parameters with respect to the ego vehicle frame.
|V. Šimundić, D. Mihelčić, D. Svirac, P. Đurović, R. Cupec (Faculty of Electrical Engineering, Computer Science and Information Technology, Osijek, Croatia)
Safety System for Industrial Robots Based on Human Detection Using an RGB-D Camera
A robotic system, in general, should be safe for the operators who are using it and, if possible, it should be safe for the system itself and the environment it interacts with. This paper investigates three methods for human detection by an RGB-D camera and their adjustments for monitored robot movement. The methods are implemented in ROS and integrated into a robotic system consisting of an industrial robot arm ABB IRB-2400L and an RGB-D camera. Validation and comparison of the implemented algorithms was performed on the recorded video dataset containing people in the robot working area with per-frame annotations of human presence, velocity and proximity to the robot. The algorithms detect humans with highest recall of 98%.
|D. Dlaka, D. Chudy (Klinička Bolnica Dubrava, Zagreb, Croatia), B. Jerbić (Sveučilište u Zagrebu, Fakultet strojarstva i brodogradnje, Zagreb, Croatia), A. Kaštelančić, M. Raguž (Klinička bolnica Dubrava, Zagreb, Croatia)
Robot-assisted Stereotactic and Spinal Neurosurgery: a Review of Literature
Development of stereotactic neurosurgery occurred due to the special area of interest within the neurosurgery, the necessity for precise targeting within the skull, and the specificity of the technique demanding superior precision. Number of technological developments and improvements, especially in pre and intraoperative imaging and guidance, as well as microscopy brought neurosurgeons to their skill limits. Thus, robotic-assisted surgery provided better posture, visualization, precision, localization and accuracy. In recent decades, robotic neurosurgery and robot-assisted neurosurgery have been developed and strengthened within it, which, in addition to the advantages of an experienced clinician, benefits from significant advantages of robots such as better preoperative planning, non-fatigue, steady motion with limitless reproducibility, superior precision, localization, and accuracy, etc. The growing area expands to spinal neurosurgery where it is mostly used for precise screw placement reducing length of the procedure, complications, and radiation exposure. Hereby, we present a review of robotic systems used nowadays in stereotactic and spinal neurosurgery.
|M. Švaco (Faculty of Mechanical Engineering and Naval Architecture, Zagreb, Croatia), A. Kaštelančić, D. Dlaka (Dubrava University Hospital, Zagreb, Croatia), F. Šuligoj, M. Marijić (Faculty of Mechanical Engineering and Naval Architecture, Zagreb, Croatia), D. Chudy (Dubrava University Hospital, Zagreb, Croatia), B. Jerbić (Faculty of Mechanical Engineering and Naval Architecture, Zagreb, Croatia)
Workspace Analysis of Robotically Assisted Percutaneous Radiofrequency Thermocoagulation for the Trigeminal Neuralgia
Trigeminal neuralgia is a sharp and paroxysmal pain that spreads to one or more branches of the trigeminal nerve and most commonly affects one side of the face. Trigeminal neuralgia can be treated by percutaneous thermoablation (radiofrequency thermocoagulation), the success of which depends on the experience of the surgeon. Percutaneous thermoablation requires high accuracy while targeting the trigeminal nerve in the foramen ovale, which is located at the base of the skull, and whose dimensions average 4x6 mm. This paper presents an initial overview on the possibilities of the application of different off-the-shelf robots in the operative procedure for treating trigeminal neuralgia. The paper simulates the reachability and applicability of the Kuka KR6 Agilus R900, KR6 Agilus R1100 and the KUKA IIWA R820 robots with 6 and 7 degrees of freedom. The operating procedure is simulated in a virtual operating room and all critical positions are validated. Possibilities of positioning the robots with respect to the patient and the recommendations for designing tool guides for performing the percutaneous thermoablation for trigeminal neuralgia are given.
Bojan Jerbić (Croatia), Marko Švaco (Croatia)
Tadej Bajd (Slovenia), Ante Bakić (Croatia), Ricardo Branco (Portugal), Bojan Jerbić (Croatia), Zlatko Katalenić (Slovenia), Igor Kotenko (Russia), Zdenko Kovačić (Croatia), Danica Kragić Jensfelt (Sweden), Jadran Lenarčič (Slovenia), Duc Truong Pham (UK), Vincenzo Piuri (Italy), Ioan Sacala (Romania), Bruno Siciliano (Italy), Karolj Skala (Croatia), Uroš Janez Stanič (Slovenia), Zorislav Šojat (Croatia)
Registration / Fees:
The discount doesn't apply to PhD students.
REGISTRATION / FEES
|Price in EUR
Up to 13 September 2021
From 14 September 2021
|Members of MIPRO and IEEE
|Students (undergraduate and graduate), primary and secondary school teachers
University of Zagreb
Faculty of Mechanical Engineering and Naval Architecture
Ivana Lucica 5
HR-10000 Zagreb, Croatia
Phone: +385 1 6168 356
Accepted papers will be published in the ISSN registered conference proceedings. Presented papers in English will be submitted for inclusion in the IEEE Xplore Digital Library.
There is a possibility that the selected scientific papers with some further modification and refinement are being published in the Journal of Computing and Information Technology (CIT).
Opatija, with its 170-year-old tourism tradition, is the leading seaside resort of the Eastern Adriatic and one of the most famous tourist destinations on the Mediterranean. With its aristocratic architecture and style, Opatija has been attracting artists, kings, politicians, scientists, sportsmen, as well as business people, bankers and managers for more than 170 years.
The tourist offer in Opatija includes a vast number of hotels, excellent restaurants, entertainment venues, art festivals, superb modern and classical music concerts, beaches and swimming pools – this city satisfies all wishes and demands.
Opatija, the Queen of the Adriatic, is also one of the most prominent congress cities in the Mediterranean, particularly important for its ICT conventions, one of which is MIPRO, which has been held in Opatija since 1979, and has attracted more than a thousand participants from over forty countries. These conventions promote Opatija as one of the most desirable technological, business, educational and scientific centers in South-eastern Europe and the European Union in general.
For more details, please visit www.opatija.hr and visitopatija.com.