PhD Work - Research Proposal -6G communications for Smart Cities

Abstract 

We are attending a paradigm shift in the requirements and solutions adopted for cities regarding mobility, connectivity, and security. These changes are driven by implementing 5G communications, developing autonomous vehicles, and Smart Cities projects. Smart cities (SC) have needs in terms of traffic control, more intelligent, integrated, comfortable, safe, and fast transport systems. Traffic control systems are expected to allow higher and secure traffic flow and efficient vehicle communications, with other equipment and people sharing the lane (V2X).Vehicles must communicate with each other(V2V), communicate with the infrastructure (V2I), and the vehicle’s position in the city context is needed and so must be known. In this scenario, the need arises to study new communication models to ensure low latency, high bandwidth, and accurate positioning information. The new communication models require the research community’s interdisciplinary and concerted effort. This project thesis proposal aims to contribute to presenting innovative proposals in the context of communications in Smart Cities. In this context, we intend to study the particular case of road crossings to propose communications for a city model with autonomous vehicles. It is intended to evaluate a model that is automatically controlled (without traffic lights), safe for pedestrians, cyclists, and autonomous vehicles, and which allows the collection of information to evaluate traffic flow in a broader context of the entire city it operates.

Introduction

Traffic flow is always considered the main objective in cities, as this reduces the time in traffic, reduces fuel consumption and pollution, and in general, it will improve the quality of life of those who use the city to work, study, or live [8]. In cities, intersections are the most complex points in terms of traffic and also the ones that cause more road accidents. The impacts of traffic on the safety of drivers and pedestrians are a strong reason for investigating safe means that can help vehicles, drivers, and pedestrians have safe and fluid access on the roads [9]. Today, the available technology already provides us with sensors, video cameras, LiDAR Radar and GPS that are installed in autonomous vehicles(AV). Those sensors can be coupled to the road infrastructure to provide systems that complement each other in the solutions and provide a good description of the environment surrounding the road, the vehicles, pedestrians, and obstacles that are found [9].

Author: Manuel Robalinho

More in this link.

PhD Work-Cooperative robotic systems for forest inventory

 Abstract

This work addresses the application of cooperative robotic systems in forest inventory task. The specific context of the use of robotic systems in forest brings a set of technical, geographic and environment constraints that shape the approach and possible techniques to be applied. In this work, we will only address the task of collecting images at selected points in the forest using a team of cooperative robots. Our goal will be, given a predefined number of robots and a predefined number of selected targets, to use an algorithm that maximizes the efficiency of the cooperative robots for the execution of the inventory tasks. Other related activities, such as the analysis and decision of which targets should be inventoried, data communication, database and volume calculations based on measurements, will be addressed in related future works.

INTRODUCTION

This work proposes an approach for the optimization of the forest inventory tasks to be performed by multiple robots, which will perform the tasks of measuring and evaluating the volumes of wood in the areas that are selected for inventory. This work is organized as follows: in section 1, we will explain the environment and applicability where this study is intended to be inserted; we will have a subsection for historical analysis and concepts of applicability of the theme. Section 2 covers related works on multi robot systems. Section 3 briefly introduces two well-known algorithms for the optimization of task assignment to robots: the Hungarian algorithm and the Vickrey auction. In section 4, we discuss and evaluate those algorithms to understand which one best suits the objective of the forest inventory. Section 5 presents the main conclusions of this work.

Author: Manuel Robalinho

More in this link.

PhD Work - Multispectral Image Compression

 Abstract

Visual experience is the main way humans communicate with each other. It is one of the main ways of interacting with our society, of obtaining and transmitting ideas, of molding projects and proof of various events and phenomena. Compressing images will always remain a research challenge, given the large amounts of information we produce in images and given the large number of devices prepared to capture them daily. Multispectral and hyperspectral image compression has received considerable interest in recent years, given the increase in image capture systems and the high resolution that imagers can easily capture, thus obtaining images with a lot of highly correlated data. The development of new areas of interest in the exploration of images also brought new contributions to the research, such as: autonomous vehicles, precision agriculture, space images, satellites communication and the medical field. This process generates very large images, which create other problems, their transmission, storage and computational power for the analysis of the data obtained. With that scenario, digital image compression has become a major factor in the operation of many day-to-day technologies. In this article, we address the issue of multispectral image compression and the algorithms used in this process.

Author: Manuel Robalinho

More in this link.

PhD work-Article Review Pseudo-LiDAR for Image-Based 3D Object Detection

Abstract

Most vehicle manufacturers and several technology companies have been studying concepts and technologies that can allow autonomous, safe and affordable driving for users of these systems. Three lines of investigation have directed the projects of autonomous vehicles, with regard to object detection: a) Using Lidar technology; b) Using stereo video cameras; c) Using both technologies in parallel. In this article, we cover the process described in the article ”End-to-End Pseudo-LiDAR for Image-Based 3D Object Detection” which combines image capture with stereo cameras to obtain a digital image of points that surround the detected objects, for the system to simulate a Lidar image, which can then be exploited with the detection and classification algorithms available for Lidar technology. With this training process, the article presents a Pseudo Lidar.

Author: Manuel Robalinho

More in this link.

PhD Work - Book Review: Robotics, Vision and Control

Abstract

The book describes the concepts of robotics and their practical application, and the concepts of computer vision applied to robotics, which involves the application of computational algorithms to data. The author has created two MATLAB® toolboxes, and explains in the book the concepts of robotics, using MATLAB®  code in all situations, thus making the book an excellent material for the study and practice of robotics. The topics covered are oriented by real problems, with many illustrative images, which makes reading accessible and didactic. Robotics and machine vision involve the application of computational techniques and algorithms for data analysis. Advances in the use of robots and machine vision in all areas of society continue to be driven by more efficient methods of processing data. The best performance doesn’t always come only from the best machines. The use of new algorithms, combined with greater computational power and cloud computing techniques, led us to the almost omnipresent participation of robots in our daily lives. It is in industry, medicine, air transport, autonomous vehicles, mining and space exploration. This article is a literature review based on Peter Corke’s book, and the application of his techniques with a use case in autonomous vehicles.

Author: Manuel Robalinho

More in this link.

PhD Work - The challenge of new communications for Smart Cities

Abstract 

We are attending a paradigm shift in the requirements and solutions adopted for cities regarding mobility, connectivity, and security. These changes are driven by the implementation of 5G communications, the development of autonomous vehicles, and Smart Cities projects. Smart cities (SC) have needs in terms of integrated traffic control and rapid transport systems. Traffic control systems are expected to enable road safety and communicate efficiently with vehicles, other equipment, and people sharing the road (V2X). Vehicles must communicate with each other(V2V), communicate with the infrastructure (V2I), and the vehicle’s position in the city context must be known. In this scenario, the need arises to study new communication models to ensure more latency, bandwidth, and positioning information. The new communication models require an interdisciplinary and concerted effort from the research community. This paper aims to contribute to the presentation of innovative proposals in the context of communications in SC.

Introduction

The implementation of 5G communications started in 2020, and the three principal axes of innovation for this communication are improved mobile broadband (MBB), reliable and low-latency communications (RLLC), and massive machine-type communications (MTC). Still, at an early stage of implementation, it is already clear that new technological advances in communications (in addition to 5G) will be necessary to obtain a capable response to the demands of society, which is increasingly digitized and more global. The quest for higher data rates, more spectrum, and more significant network densification means driving the future of wireless services to provide at least a terabit-per-second aggregate bit rate in small regions [1]. A new cycle of communication development began with

implementing 5G technology. Over approximately ten years of studies for implementing 5G, many other needs emerged, new use cases, new equipment, and materials. Society has realized that a new impulse in communications will be necessary to guarantee the dreamed features. As usual, each technological cycle has been implemented in approximately ten years, so it is expected that 6G will provide a solution to the use cases presented by society today and be implemented around 2030.

One of the unavoidable issues in the discussion of SC is traffic management, especially for private vehicles.

More in the link.

Author: Manuel Robalinho and my coordinator Dr. Pedro Fernandes