A Wearable Embedded System for Health Data Transmission and Patient Real Time Spotting

Transcription

1 A Wearable Embedded System for Health Data Transmission and Patient Real Time Spotting A. Bozinaki, E. Printezi, M. Papoutsidakis and D. Tseles During the length of this research, we conceived and developed the idea of an automatic wearable bracelet that can monitor a patient's health status. The work consists of four chapters with each one detailing the thought process and construction of the device. Firstly, there is a historical review of the evolution in medical engineering and all the milestones that lead to modern biotechnology, with an emphasis on devices worn by the user i.e wearable gadgets. Following up, we have the chapter detailing the constructional part of the project, starting with the selection of each part that the device consists of and their characteristics, to the technical design and how it brings all the parts in a cohesive working machine. This chapter is called hardware analysis. The next chapter analyzes the gadget's function scenario as well as the code written for the device to meet its requirements. We begin with the software analysis by finalizing the variables, followed by the once-occurring function in which some required installations take place, then the basic program function and lastly all the required functions that were designed to complement the initial code. As technology is constantly evolving, it is only logical that there is always room for improvements and upgrades. The fourth chapter details all the conclusions made and every problem that occurred during the idea's implementation, as well as observations and ideas that can improve and make the bracelet's function more sophisticated.

2 Disrupting the spare parts supply chain in the shipping industry with the introduction of 3D printing E. Kostidi, N. Nikitakos 2 Mechanical Engineer, Postgraduate Student at the Department of Shipping, University of the Aegean, Address: Lesvou 0,8200, Chios, Greece, cellphone: , evanthiakostidi@gmail.com 2 Professor at the Department of Shipping, University of the Aegean, nnik@aegean.gr Additive manufacturing (or 3d printing as is commonly known) has been already implemented in various sectors (industrial products, consumer products, automotive, aerospace, etc.).the main benefit of this technology is that it allows flexible production of customized products at no extra cost in manufacturing. This is achieved by directly converting the 3D data into physical objects, without the need for additional tools or molds. Furthermore, the principle of the construction in layers can produce functionally integrated components in a single production step, thereby obviating the need for the assembly stage. The aim of this work is to study the disruption in the supply chain of the spare parts of the ships with the introduction of additive manufacturing (3d printing). A literature review of the available technology (methods, materials), and the benefits and impact of its application, especially in industries with similar to shipping characteristics (industries with moving assets), reveals the potential of applying it in the shipping industry. The as-is process is modelled including the involved stakeholders (the ship, the land office, the suppliers, the manufacturer etc.) in order to get an understanding of it. A to-be scenario is proposed pointing out locations (decentralised near the point of demand) for 3D printing spare parts in order to get the benefits (i.e. response time, reducing inventory, minimise transportation cost). Recommendations for further research are made. Keywords: additive manufacturing,3d printing, supply chain, spare parts, shipping industry. 2

3 A Comparison of Classical and Intelligent Control Methods of the I-Term Effect in Hydraulic Systems P. Papageorgiou and M.Papoutsidakis Piraeus University of Applied Sciences (PUAS) Dept. of Automation Engineering Athens, Greece With the continuous development in industrial machinery, a higher demand for system stability and accuracy occurs. Hydraulic systems consists of several dynamic parts which are widely used in motion control applications. These dynamic parts need to be controlled to determine the direction of the motion. The purpose of this paper is to evaluate the performance of the hydraulic actuation system by using a different number of control methods, in Matlab / Simulink environment with Sim Hydraulics. This process began by modeling the system. The system is divided in two portions, a hydraulic and a mechanical. Classic PID controller and fuzzy PID controller are designed. The design and implementation of a self-tuning-parameter fuzzy PID controller is described alone with controllers rule structure. Based on the input and output data of the system, comparisons are made. By analyzing the results obtained through this experimental work, conclusions are made from the usage of the I-term and its effect on the system s behavior. Keywords: PI, conventional PID, fuzzy PID controller 3

4 Simultaneous Localization and Mapping for micro Unmanned Aerial Vehicles P. Tsilivis, G. Nikolaou 2 Dpt. of Automation, Piraeus University of Applied Sciences, Athens, Greece Tel: , p.tsilivis0@gmail.com 2 Dpt. of Automation, Piraeus University of Applied Sciences, Athens. Greece Tel: , nikolaou@teipir.gr The aim of this paper is to present and analyze the basics of the simultaneous localization and mapping (SLAM), implemented in micro Unmanned Aerial Vehicles UAVs. In particular, one of the most essential issues while developing a project to navigate a drone autonomously, is the localization and mapping. For a UAV, the ability to predict its position and recognize the surrounding environment, is of the utmost importance. Hence, there are numerous techniques that rely on different parts of a drone, such as navigation based on camera or localization based on GPS and laser beacons; these methods improve the performance of the self-navigating drone regarding its stability during the flight and its response in unknown environments. Owing to the advanced level of difficulty, these algorithms were developed with the view of operational consistency, not for perfection. Thus, this paper will initially present the fundamentals of SLAM, along with a detailed description of the importance of SLAM algorithms in regards to the autonomous navigation of a UAV. Secondly, the most essential and commonly used SLAM algorithms will be descripted theoretically in-depth, based on existing researches. In the last part of this paper, a practical application will be presented based on a Parrot AR. Drone 2.0. along with various tests about the stability of the AR Drone. Finally, the most crucial challenges of the autonomous navigation will be descripted, based on the particular drone. Keywords: SLAM, micro-uav, Autonomous Navigation, Localization 4

5 Short Text Classification Application in Automated Workflow Management Systems Iv.Javakhishvili Tbilisi State University Manana Khachidze, Magda Tsintsadze, Maia Archaudze, Gela Besiashvili {manana.khachidze, magda.tsintsadze, maia.archuadze, Despite abundance of various Workflow Management Software Products, many companies prefer to develop own software according to their specific requirements, especially for Automated Workflow Management Systems. The presented work considers the mechanism of document processing on the example of one of the largest educational institution of Georgia, based on NLP (Natural Language Process) methods. The classification process of Georgian language based texts is described. Each step of document workflow management starting from text-initial processing to machine learning is provided. The realization algorithms of every issue consider Georgian Language peculiarities. Key words: Text classification, Workflow, NLP, Machine learning, Text processing 5

6 Dental Self-diagnostic Information System Based on the Natural Language Processing Iv.Javakhishvili Tbilisi State University Maia Archaudze, Gela Besiashvili, Nia Khachidze, David Khachidze { maia.archuadze, gela.besiashvili}@tsu.ge, nia_khacho@yahoo.com, khacho_dg@yahoo.com Nowadays information systems are actively used in Dentistry, as well as, in general medicine. Systems with their purposes are diverse, but self-diagnostic systems are less represented. Despite the fact that self-diagnostic systems have number of downsides they should not be neglected. The paper presents basic scheme for the Dental selfdiagnostic information system. Self-diagnostic procedure is based on patients description of their condition by means of natural language texting. There are presented different key words for describing variety of diseases. These key words are so-called Hints. By using Hints patients describe their condition stepwise. After analyzing the submitted information on each step, patients are provided with new cycle of Hints and so on. The procedure of diagnosis is based on Decision tree algorithm. Multitude of Hints is obtained by method of Info Gain characteristics selection. Key word: Self-diagnostic Information System, NLP, machine learning, Decision tree, Information gain. 6

7 Predicting Environmental Data in Public Management by Using Artificial Intelligence Georgios N. Kouziokas, Alexandros Chatzigeorgiou 2, Konstantinos Perakis 3 Dpt. of Planning and Regional Development, School of Engineering, University of Thessaly, Volos, Greece gekouzio@uth.gr 2 Dpt. of Applied Informatics, University of Macedonia, Thessaloniki, Greece achat@uom.gr 3 Dpt. of Planning and Regional Development, School of Engineering, University of Thessaly, Volos, Greece perakis@uth.gr This paper focuses on building neural network models in order to predict environmental information that is crucial in environmental decision making in public management and planning. The application of artificial intelligence in various fields have been highly increased the last decades with the development of new neural network learning techniques and tools in constructing neural network models. In this research, the application of Feedforward Neural Network Models for predicting environmental data is implemented. Environmental data can play an important role in public administration and also in environmental management and planning by promoting sustainable decision making strategies. The levels of pollution factors were chosen to be predicted, since pollution factors are associated with public health problems which is of high importance. Several artificial neural network models were constructing by using different architectures regarding the number of the neurons in hidden layers, the number of the hidden layers and the input and output neurons in order to build the optimal model that would predict efficiently the environmental data. Multilayer Feedforward Perceptron was used in this research as it is the most suitable for time series forecasting according to many researchers. The forecasting model can be valuable for public administration, since it can be used as a tool for a more efficient environmental management and also in adopting proactive measures and policies. Keywords: Artificial intelligence; Neural networks; Environmental Planning and Management; Public Management. 7

8 An Information System for Monitoring Environmental Indicators in Public Management for Sustainable Development Georgios N. Kouziokas Dpt. of Planning and Regional Development, School of Engineering, University of Thessaly, Volos, Greece This paper presents an information system that was developed based on monitoring the environmental indicators in order to provide a decision making framework in public management for sustainable development. The developed system can be used to store and elaborate information regarding predefined environmental indicators of many kinds regarding soil, water and air analysis (CO, NO, etc.), facilitating decision making in public administration. The interpretation of the variances of these factors can help stakeholders to build predictive models and perform risk assessment for the conservation of the natural environment and also the for the public health. This paper describes the software application, its structure and the way that can be utilized in public administration in order to take decisions and proactive measures regarding environmental issues and to apply the adequate environmental policies. Considering the development of Information and Communication Technology (ICT) and the highly increased amount of environmental data, it is significant for the authorities to utilize an information system that will manage all relevant environmental information and in order to assess the environmental dangers in a more holistic way by using all the collected data and the environmental indicators as a tool for a more efficient environmental and urban management. Keywords: Environmental Information System; environmental indicators; Environmental Planning and Management; Public Management. 8