“e-health Sri Lanka”- an IoT Readiness in Health Sector”

Abstract— Traditional methods of physical contact with the patients are expanding to the virtual platform of e-health in modern healthcare services. Computer based healthcare systems are now serving the societies around the planet. E-health is not simply hardware or software, it includes health management information. This paper extracts and colligates the determinants of e-health using research done for other countries to determine the perspectives of e-health in Sri Lanka. An analytical study was done to find out the impression of patients in Sri Lanka in using IoT in health sector in Sri Lanka.

Keywords—IoT, e-health,HealthMIS

Authors-P.D.Suranjini Silva Department of Information Systems Engineering Sri Lanka Institute of Information Technology Malabe,Sri Lanka suranjini.s@sliit.lk , J.A.D.C.A. Jayakody Department of Information Systems Engineering Sri Lanka Institute of Information Technology Malabe,Sri Lanka anuradha.j@sliit.lk

View Full Paper

Extracting Unstructured Data and Analysis and Prediction of Financial Event Modeling

Abstract—The stock market is full of uncertainty. It always tends to change with internal and external factors.  Since it’s commonly accepted that the failure in the Effective market hypothesis, in this research try to focus on the few external factors. Stock price movements are driven by publications of financial news. Investors’ or sellers’ decisions are made based on the available information considering how a stock market will going up or going down for next few days. News articles incorporate information about a firm’s fundamentals, the activities in which a firm is involved and the expectations of other market participants about future price changes. This research tries to focus on the few external factors. Predicting the future expecting variations in the stock market can be by analyzing. This analysis can be done on three types. Those are fundamental, technical and machine learning. In the research paper use the technical analysis. NLP use to Data extraction in the means of a summery, from a selected website. Then clustering and classification into external factor using K-means and KNN. The impact of the extracted summery for the stock market prediction is analyzed using sentiment analysis. Here it will analyses how each external factor implies on the selected sectors in the stock market, therefore it can help in the prediction of actual stock movement for the sectors. Prediction based on Naïve Bayes, Recurrent Neural Network (RNN) approach and the sentiment score of the text segment is calculated by sentiment analysis of the text through SVM.

Keywords- Naïve Bayes, RNN, K-Means, KNN, NLP, SVM, sentimental analysis, clustering, classification, stock market prediction

Authors- B.Tharaniya Faculty of Computing Sri Lanka Institute of Information Technology Sri Lanka tharaniyabala@gmail.com, Lakmal Rupasinghe Faculty of Computing Sri Lanka Institute of Information Technology Sri Lanka lakmal.r@sliit.lk , Chethana Liyanapathirana Faculty of Computing Sri Lanka Institute of Information Technology Sri Lanka chethana.l@sliit.lk , Kalpa Kalhara Sampath Faculty of Computing Sri Lanka Institute of Information Technology Sri Lanka kalpa.s@sliit.lk

View Full Paper

Development of a Two-Dimensional Plotter

Abstract – Two-dimensional plotter concept has been used in many applications since a long time. Plotters are controlled by computer. Plotter is a device which do drawings according to the given coordinates. Most popular XY plotter designs are two axis manipulators, stepper motors included, and timing belt driven mechanisms are included. Main applications of plotter are CNC machine, Engraving machine and drawing machine etc. XY plotter operates in two perpendicular axes of motion in order to provide two independent movements and drawings. These type of plotting mechanisms normally use stepper motors as it is easy to achieve accuracy unlike DC motors. In drawing applications, usually image processing tools are involved to perform edge detection, noise removal and mapping etc. There are image processing software available which supports advanced processing methods such as OpenCV, Matlab. Different types of plotters are being used commonly such as cutting plotters, XY writers, electrostatic plotters, drawing machines etc.

Keywords – Plotter; Robotics; Drawing Robot; Image Processing

Authors- N. Gallege School of Engineering and Technology Asian Institute of Technology Bangkok, Thailand  tilanigallege@gmail.com , Manukid Parnichkun  School of Engineering and Technology  Asian Institute of Technology  Bangkok, Thailand manukid@ait.asia

View Full Paper

Formulation of a National Framework for the Digital Transformation of Sri Lanka’s Public Bus Transport System

Abstract – Public transport play an increasingly important role in human navigation. In a digital world where the technological landscape continues to change, the need for a quality transportation network is critical. Sri Lanka being a developing country has witnessed digital transformations in several sectors although the use of digital practices in Sri Lanka’s bus transportation industry is minimal. The proposed research aims to present how IT can be used to deliver a quality transportation experience to each passenger through the use of a real-time bus tracking system that uses GPS coordinates along with a bus arrival time prediction system capable of estimating the arrival time of the desired bus. The framework also aims to introduce a cashless ticketing mechanism to eliminate the problem of carrying change money which is a burden faced by many passengers. When real-time information is available, the passengers could spend their time efficiently and reach the bus stop on time or take alternate paths if the bus is to be delayed. This in return would be capable of solving the problem of traffic congestion as the availability of a high-quality bus transport service would encourage citizens to use public transport instead of their own private vehicles.

Keywords-GPS; digitalization; bus transport; traffic congestion; passengers

Authors- R. Dunuwila, J.A.D.C.A. Jayakody, R.T. Weerawardhana ranshari@gmail.com, anuradha.j@sliit.lk, rashmeetharindra95@gmail.com , W.A.S. Pabasara, U.S. Ketipearachchi, H.M. Sunas Department of Information Management Sri Lanka Institute of Information Technology Malabe, Sri Lanka weerasinghesamadhip@gmail.com, udarashalini@gmail.com, sunasmhd@gmail.com

View Full Paper

Concepts to Maximize the Usage of Distribution Networks – Case study on Jaffna Peninsula Distribution System

Abstract – Usage of distribution networks are always a challenging task as it directly connected to the consumer loads, which are continuously varying. Under or over voltages are one of the major component that limits the expansion of electrical distribution system. This can be eliminated by operating all devices to their optimize positions. This done through applied research study in this paper, which resulted a concept on operational steps while giving a cost effective solution in overall. Here the PSCAD Bergeron model is used for modelling the Jaffna peninsula electrical distribution system. The operation is studied for the full day load pattern thus showed voltage problems in some bus bars especially during peak load conditions. Solutions are proposed and healthy operation is validated with proposed solution. Further the concept on finding the solution is also concluded a general contribution from this study.

KeywordsDistribution system study, under voltage, concept on distribution system operation, Jaffna peninsula network and distribution system Bergeron model

Authors- Ravindran University College of Jaffna Sri Lanka ravimathy@yahoo.com , W.D.A.S. Rodrigo, W.D. Prasad University of Moratuwa Sri Lanka asankar@uom.lk, prasadwd@uom.lk, Atputharajah University of Jaffna Sri Lanka atpu@eng.jfn.ac.lk

View Full Paper

IoT Based Intelligent Railway Level Crossing System

Abstract –Railway level crossing management has become a critical need as there is an increased number of accidents in unprotected railway gates in Sri Lanka. The goal of this project is to identify a dynamic, reliable and cost effective solution for a railway level crossing system which not only operates in a fully automated manner, but also can be remotely controlled by a central control room on demand. Though railway gate management systems exist in Sri Lanka, the cost of managing such systems is much higher as it requires human resource in order to perform manual tasks.IoT is an emerging technology which allows the physical world to be sensed through sensors, respond and take actions through actuators. An IoT solution for such problem would be ideal as it manages the system effectively and considering its ability to improve itself over time using the data gathered by the system.The main functions of the proposed system are to detect the train from a defined distance, operate the gates automatically and to warn the vehicles when they are trying to enter the unsafe area while the gates are closed. A web dash board is provided to the central control room for monitoring the events from each gate. If manual operation of any gate is required, the central control room can perform it using the controls on web dash board. The system consists of vibration sensors to detect the train, servo motors to operate the gates, ultrasonic sensors to detect the vehicles, buzzers and LEDs to warn the vehicles which enter the unsafe area. The main compute module is Raspberry Pi 3 Model B. These actions are performed and the effectiveness of the system is tested using a prototype which is built under laboratory conditions. 

Keywords-Railway level crossing, IoT, MQTT, Raspberry Pi 3 Model B

Authors-Fernando W I L R Faculty of Graduate Studies and Research Sri Lanka Institute of Information Technology Colombo, Sri Lanka lahiruf137@gmail.com ,  Jayawickrama J A D T Faculty of Graduate Studies and Research Sri Lanka Institute of Information Technology Colombo, Sri Lanka dinuwanadg@gmail.com, Jayasiri D D Faculty of Graduate Studies and Research Sri Lanka Institute of Information Technology Colombo, Sri Lanka dushani@geoidit.com, Somarathna V Faculty of Graduate Studies and Research Sri Lanka Institute of Information Technology Colombo, Sri Lanka vimuths2@gmail.com

View Full Paper

A Framework for Dynamic Pricing Electricity Consumption Patterns via Time Series Clustering of Consumer Demand

Abstract –A time series is a quantity sampled regularly in time where the techniques used in its modeling and characterization are dependent on the complexity of its statistical properties. Domestic electricity consumption exhibits time series in the form of time varying usage patterns. These patterns are of fundamental importance in determining load based pricing schemes and in more effective generation and distribution planning. Clustering individual household electricity consumption patterns enables a utility to design pricing plans catered to groups of households in a particular locality to more accurately reflect the cost of supply at a particular time of day. Dynamic pricing is an attempt to change the consumption behavior to one that is more uniform and devoid of sharp peaks ensuring a more uniform utilization of generation capability avoiding idle capacity at times of low demand. In light of this we model the movement of collections of such time series in order to observe the relative movements between them with a view to understanding their group behavior and characteristics not otherwise outwardly visible. Time series clustering can be accomplished by directly clustering samples of the series using an established method of clustering or one could estimate various representative characteristics of each series that are then used to parameterize the series and form the attributes used in one of the standard clustering techniques. In this paper we model each time series as an Autoregressive Moving Average (ARMA) process with an optimal model order determined by the Akaike Information Criterion when the parameters estimated by the Hannan-Rissanen algorithm converge. The estimated model has the representation of a transfer function with a frequency response defined by the ARMA parameters. We use the frequency response as the means to further refine the within cluster profiling and classification of the objects. Through our modeling we are also able to identify instances where the consumption behavior exhibits patterns that are uncharacteristic or not in line with the behavior or consumption profiles of the other households in a particular locality providing insights in to potential faults, fraud or illegal activity. 

Keywords– Time Series, ARMA, Clustering, Stochastic Processes

AuthorsAsoka Korale, Ph.D., C.Eng., MIET Surveillance Department Millennium Information Technology Colombo, Sri Lanka  akorale@sltnet.lk

View Full Paper

Smart BPM from the Perspective of a Telecom Service Provider for Incident Management

Abstract –Business Process Management (BPM) is in the throws of an extreme makeover. The traditional focuses of process efficiency and cost reduction have given way to agility and digital automation. For Telecom service providers, BPM place a premium on evaluating and providing leading-edge capabilities. With that, the role of Artificial Intelligence (AI) has come to the forefront as a mean to drive better and deeper customer experiences within a digital process. While BPM drivers are shifting toward digital automation and customers, AI is emerging as a means of better serving customers. AI’s ability to take human cost and latency out of processes, as well as provide new interfaces that delight users, mirrors the new goals of BPM. This paper examines the current state of BPM and planned AI improvements among leading telecom business processes.

Keywords– Business process management, Artificial Intelligence, business process implementation

Authors- Sakisan S Engineer-BPR Sri Lanka Teleocm PLC sakisans@slt.com.lk , D R N Jayasuriya Deputy General Manager-BPR Sri Lanka Teleocm PLC naleen@slt.com.lk

View Full Paper

A Comprehensive Analysis on Social Engineering based Attacks

Abstract –This research presents a comprehensive study on social engineering attacks. An introduction of social engineering attack is given, with context to the current trends and related vulnerabilities which comprises of all aspects related to Social Engineering, thus giving a defined overview of what social engineering is and how it is affecting us all. The main reasons for the spread of social engineering attacks in the current context are discussed with the aid of an attack framework and a proposed defense in depth approaches for social engineering threats. Further, this paper presents concludes the remedies which can be taken as a safety precaution for attacks and social engineering

Keywords– attack framework model, defense approaches, social engineering.

Authors- D.S.A.Kandawala Department of Research and Postgraduate Sri Lanka Institute of Information Technology Malabe, Sri Lanka. sachini.k@sliit.lk , Anuradha Jayakody Department of Information Systems Engineering Sri Lanka Institute of Information Technology Malabe, Sri Lanka. anuradha.j@sliit.lk , J.K.S.Roledene Department of Research and Postgraduate Sri Lanka Institute of Information Technology Malabe, Sri Lanka. sroledenez@gmail.com

View Full Paper

Solar Powered Automatic LED Lamp System with Power Optimization for Southern Expressway

Abstract –The best solution for reducing power generation cost during peak hours is, shift some loads within peak hour to off peak hours. This solution cannot be subjected to light load. So ideal solution for removing light load from peak hours is to isolate those from grid during peak hours. Street lamps are having a considerable light load depend on grid. If we can remove some of this street light load from grid it would be a great advantage. In this paper we propose to convert all street lamps in Southern expressway to solar power LED lamps. Besides reducing daily peak load, this will provide several other advantages as well. The system is environmental friendly, cost effective with only installment cost, few or no maintenance, long life than existing system and high efficient. Solar electricity is free. Because of these notable advantages the proposed system can be used to replace existing street lamps. Proposed solar lamps can be installed in the extended part from Matara to Hambantota as well.

Keywordscomponent; formatting; style; styling; insert (key words)

Authors- Anurudde Nirmala Nanayakkara Sri Lanka German Railway Technical Training Center Sri Lanka Railways Ratmalana, Sri Lanka  annirmala@gmail.com

View Full Paper

BringOWN – An Agile BYOD Solution with Device Friendly Ciphering

Abstract –Recently, BYOD or Bring Your Own Device has become one of the most popular model for enterprises to provide mobility and flexibility in workplaces. The emergence of new technologies and features of mobile devices makes them integral parts of every aspect of daily business activities. In BYOD, the personal devices can be used to increase employees’ satisfaction and reduce an organization’s device costs. However, due to attacks and vulnerabilities it is difficult to trust the personal devices coming into the workplace. There is security concern to protect sensitive corporate data, protecting sensitive corporate data is a major concern in the industry. This project proposal proposes a BYOD model which can be used to overcome the mention problems/vulnerabilities. The BYOD model is a combination of a light weight

KeywordsBYOD;Android;Software Security;VPN;FTP; Encryption.

Authors- Shashika Lokuliyana1, Suranjini Silva2, Warnakulasuriya K. T. T3, Dilan Bandara4, Peshala Vithana5, Sahan ShavindrikaDepartment of Information Systems Engineering, Faculty of Computing Sri Lanka Institute of Information Technology, Malabe, Sri Lanka 1shashika.l@sliit.lk, 2suranjni.s@sliit.lk, 3krishees@yahoo.com, 4dilanchathurangamlz@gmail.com, 5peshalawith@gmail.com

View Full Paper

An IoT Based Hardware Filter for Internal Network Security

Abstract –Establishing internal network security of IoT takes a significant importance in modern world. There were several incidents reported in relation to the attacks to the systems making critical damages and incurred significant cost to them. These attacks may cause destruction on to the lives of the people too. The most of attacks have been done remotely through injecting malware in to IoT devices or programmable logic controllers. Among these many devices are wifi enabled and firmware updatable. As a result the devices become more vulnerable.The Esp8266 is Wifi enabled microcontroller that is commonly used for low power wifi applications. It is one chip everything included and current market has many ESP8266 based devices. The automobiles, programmable logic controller (PLC) in bulding management, medical equipment, supervisory control and data acquisition (SCADA) are areas included in IoT  security at critical level.This research is focussed to find a solution for thsese issues in  IoT  security. The suggested solution is implemented through microcontroller. Research results revealed that the solution is very successful in accuracy and performance. There is an opportunity for future researchers to develop secure transmission and filtering mechanisam for any IoT  device compatible.

Keywords-Wi-Fi; IoT; Code Injection; Bitcoin; Encryption; Device vulnerability; Filter;ESP8266; Remote Access

Authors- S.A.D. Oshada Viduranga Chandrakumara. CCV Solutions  Kalutara, Sri Lanka oshada@ccvsolutions.com , Anuradha Jayakody Faculty of Computing Sri Lanka Institute of Information Technology Malabe, Sri Lanka anuradha.j@sliit.lk

View Full Paper

Improving the Basic Insulation Level (BIL) of Distribution Transformers

Abstract –Distribution transformer is one of the critical parts of the electrical distribution network to ensure reliable supply of electricity to homes and industries. During the life cycle, distribution transformer experience several electrical stresses due to lightning strikes. This results in the reduction of dielectric strength of the transformer and lifespan may reduce. For the worst case transformer will face a complete breakdown. Therefore it is important to design the transformer to meet the required insulation levels as specified in IEC based on rated voltage of the transformer or as per customer specification. A project was conducted to design, manufacture and test 33 kV rated distribution transformers with BIL of 250 kV.  Two  methods were adopted, the first by introducing extra layer insulation in HV winding and the other by inserting an electrostatic shield inside the HV winding of the  transformer. Both types of transformers were subjected to impulse withstand voltage test relevant to 250 kV BIL. The two methods were analyzed against each other by comparing the impulse test results, temperature rise test results and their material cost. In this way the most effective method was identified which has more advantage in respect to performance and material cost.

Keywords-Distribution Transformer; Electrostatic Shield; Impulse; Dielectric Stress

Authors- D. A.Gunawardhana1, R. C. Welagedara2, R. R. L. L. Wijayaratne3 ,N. W. N. J. Nanayakkara4 LTL Transformers (Pvt) Ltd, Sri Lanka 1thilini@ltl.lk,  2rasika.welagedara@ltl.lk,  3lochana.wijayaratne@ltl.lk,  4nipun.janitha@ltl.lk

View Full Paper

Intelligent Micro-Service Auto Scalar

Abstract –this research paper presents a new way of scaling Docker  micro service instances using a reactive, machine learning model based on the HTTP request load and the resource consumption of the application. The paper presents how micro service application behaves and key performance metrics which needs be consider to scale the instances and how the auto scaling is achieved using machine learning algorithms. The initial approach of the development of this project was to collect the necessary analytical data for developing the model which was used for the testing of the existing systems. It was also necessary to develop a rule based scaling system. Based on the results of the experiment it was decided that a hybrid based approach would be more efficient compared to the systems in existence. The outcome of the project is to manage the micro services with the least interaction from the human user.

Authors- Kasthurirathne1, D.P.Liyanage2, S. Ganeshalingam3, A.M.Firnas4, W.J.J.AbayarathneFaculty of Computing,Sri Lanka Institute of Information Technology, New Kandy Road,  Malabe 1dharshana.k@sliit.lk, 2it13148478@my.sliit.lk, 3it14056826@my.sliit.lk, 4it14064432@my.sliit.lk, 5it10230848@my.sliit.lk

View Full Paper

Building Smart Homes with Narrowband Power Line Communication (NB-PLC) Networks: Benefits and Challenges

Abstract –Currently Wi-Fi is the predominant technology used for building smart homes. However, Wi-Fi in this domain has several drawbacks such as loss of wireless signal, “leaky” Wi-Fi and network security issues. Although many endpoint devices (e.g. “connected” light bulbs, thermostats) and services (e.g. voice activated devices) are now available in the market for  building  smart homes, still there are no homes with all or most of the defining features of a smart home because Wi-Fi cannot, in our opinion, provide a reliable and secure home network. A smart home should provide homeowners security, comfort, convenience and energy efficiency allowing them to control smart devices, often by a smart home app on their smartphone or other network device. Unlike Wi-Fi, NB-PLC network can provide a reliable and secure command and control communication network. This paper describes NB-PLC system developed using Cypress® NB-PLC modem reference design. End point devices such as electric switches, plugs, bell push, fan controller and IR Controller with embedded NB-PLC modem along with an NB-PLC modem embedded Internet of Things (IoT) gateway have been developed and they are now being marketed. Endpoint devices do not have a directly accessible IP address. Authorized smartphones can access end point devices through the home IoT gateway. Energy monitoring and standby power management are available in some devices. NB-PLC modems can be embedded in any other electronic device (i.e. air-conditioner) easily in the future. This system would be cheaper to install and own than multiple Wi-Fi devices. It can co-exist with Broadband PLC devices (i.e. HomePlug). There are challenges in deploying this technology. NB-PLC networking for smart homes is a new concept to be proven in use. A neutral wire in the switch point is needed when current standards of house wiring do not require it. Small LED spot lights (less than 5Watt), which are increasingly used in homes today, produce electrical noise, which interferes with NB-PLC data communication. This was remedied by designing suitable filters. In addition, IoT gateway has built-inmonitoring of nodes to achieve reliable communication and message delivery

Keywords- Smart Home; Power Line Communication; Home automation; PLC modem

Authors- Jeremy Fernando MIET (UK) Managing Director / CEO. Innovative Smart Solutions (Pvt) Limited Sri Lanka jeremyF@smarthome.lk

View Full Paper

Reactive Power Control for Regulation of Voltage with Integrated Wind Farms

Abstract –Maintaining the transmission network voltage within a narrow range is required as a performance standard and also to ensure the stability of a power system. Voltage regulation is generally achieved through reactive power control. Controlling reactive power with the aid of Flexible Alternating Current Transmission (FACT) devices is a recent advent to power system operations with many a research ongoing. Further, due to the high penetration of grid connected renewable energy sources, it is becoming a trend to use the inverters associated with the renewable power generators to support voltage regulation. The operation of voltage regulating devices should comply with the variations in consumer demand as well as with the variation in power generation which is increasing becoming more stochastic due to the introduction of more and more intermittent renewable energy resources. This paper discusses a design of a reactive power control algorithm which utilizes conventional as well as inverter based reactive power support. The performance of the control algorithm when embedded in a power system simulation environment validates its ability to maintain specific voltage levels in the power system. The Sri Lankan power system is taken as the study case for all presentations in this paper.

Keywords-Voltage regulation, reactive power control, reactors, wind inverters, power intermittency

Authors- Dharmabandu1, R.W.P. Thilinika2, B.A.J. Pushpakumara3, G.D. Kanchana4, Jayatunga5, A. S. Wickramasinghe6 Department of Electrical Engineering University of Moratuwa  Sri Lanka 1charith.dharsana@gmail.com, 2prasadirathnayaka91@gmail.com, 3bajpushpakumara@gmail.com, 4dkgeeganage@gmail.com, 5upuli@uom.lk, 6amila@rmaenergy.lk

View Full Paper

Native Cloud Support for Running WSO2 Middleware on Microsoft Azure

AbstractWSO2 Carbon is the core platform on which WSO2 middleware products are built. WSO2 products can be deployed in on premise servers as well as in cloud environments such as Microsoft Azure and Amazon Web Services (AWS). Apache based WSO2 products uses membership schemes to manage the members in their clusters. Addition and removal of members in a cluster and maintaining the consistency of the members list is the main objective of a membership scheme. Well-known address based membership scheme and multicast membership schemes are in-built membership schemes supported by Apache axis2, which are most suitable for in house server clusters. Apart from them, Kubernetes and AWS membership schemes are specially developed for cloud based deployments, for the cloud platforms Google Kubernetes and AWS respectively. The authors focuses on deployment of WSO2 product clusters in Microsoft Azure including a carbon membership scheme that uses Azure native features to automatically discover members. This new membership scheme guarantees that members in a WSO2 product cluster is managed smoothly. In addition, the authors focus on development of load balancing, auto healing and centralized logging using native features provided by Azure.

Keywords-Membership Scheme, WSO2 Custer, Azure REST API, Auto Discovery, Clustering in Azure

Authors- Rathnayake O. N 1, Dumendra W.A.S 2, Jayaratne V.V.M 3, Barthelot A.J 4, Shashika LokuliyanaDepartment of Computer Systems and Networking, Sri Lanka Institute of Information Technology (SLIIT), Malabe, Sri Lanka. 1 osuranew@gmail.com, 2 sahan.wickramaarachchi@gmail.com , 3 vish120292@gmail.com , 4 allan.barthelot@gmail.com, 5shashika.l@sliit.lk

View Full Paper