Lionel Nkenyereye

A Remote System for Monitoring Auxiliary Data Center from Environmental Threats with Lower Hardware Cost

There are some data centers that are located far from the main data center for ensuring continuity of companys Information Technology operations in case the main data center encounters a serious downtime especially for companies engaged to work in distributed manner to increase quality-of -service to their customers. Environmental downtime is a significant cost to organizations and makes them unable to do business because what happens in the data center affects everyone. In addition, the amount of electrical energy consumed by data center increases with the amount of computing power installed. One strategy of reducing energy consumption in data centers is to adjust the temperature and humidity data. Installation of physical Information Technology and facilities related to environment on monitoring temperature, humidity, power, flood, smoke, air flow, room entry is the most proactive way to reduce the unnecessary costs of expensive hardware replacement or unplanned downtime and decrease energy consumed by servers. In this paper, we present remote system for monitoring auxiliary data centers implemented using open-source hardware platforms, Arduino, Raspberry Pi, and the Gobetwino. The objective of collecting temperature and humidity data allows monitoring servers health and gets alerts if things start to go wrong. When the temperature hits 50oC., the supervisor at remote headquarters would get a SMS, then take appropriate actions that make sense to reduce electrical costs and preserve functionality of servers in auxiliary data centers. The data center environmental monitoring represents a step forward towards addressing awareness monitoring of energy efficiency and data centers located far with ability to receive a notification by email.

A Study of Big Data Solution Using Hadoop to Process Connected Vehicle’s Diagnostics Data

Recently, we have witnessed a period where things are connected to the Internet; the vehicle is not leaved back because connected car field is currently explored. It is without doubt that connected vehicles will generate a huge of vehicle’s diagnostics data which will be sent to remotely servers or to vehicle’s cloud providers. As the amount of vehicle’s diagnostics data increases, the actors in automotive ecosystem will encounter difficulties to perform a real time analysis in order to simulate or to design further services according to the data gathered from the connected vehicle. In this paper, Apache Hadoop framework and its ecosystems particularly Hive, Sqoop have been deployed to process vehicle diagnostics data and delivered useful outcomes that may be used by actors in automotive ecosystem to deliver new services to car owners. A study of big data solution to process vehicle diagnostics data from connected vehicles using Hadoop is proposed.

Integration of big data for querying CAN bus data from connected car

Data transmission by Connected Car via wireless communications technologies enable new in-car telematics services. The capability to efficiently process large volume of Controller Area Network (CAN) bus data within a reasonable time. Since these data are essential for many Connected Car applications, querying and extracting useful information using Hadoop framework will allow to enhance safety and driving experience. This paper studies design steps to take in consideration when implementing MapReduce patterns to analyses CAN bus data in order to produce useful data that are hosted in the cloud. In addition, we implement a mobile apps for collecting and transferring CAN bus data to remote data center which include application server and Hadoop ecosystem such Hive data warehouse. Experiment results show that MapReduce join algorithm is highly scalable and optimized for distributed computing than Statistical Analysis System (SAS) framework and HiveQL declarative language.

Design of IoT Gateway based Event-Driven Approach for IoT related Applications

사물 인터넷(IoT)은 효율적인 시간 응답 및 처리를 위해 이벤트 중심으로 접근 할 필요가 있다. IoT에서 모바일 기기의 성장은 IoT 응용 프로그램과 관련이 있는 지능형 건물로 연결이 된다. 예를 들어, 홈 오토메이션 제어 시스템은 홈 서버에 액세스하기 위해 스마트 폰이나 웹 서비스에 클라이언트 시스템과 같은 웹 응용 프로그램을 사용하여 제어 명령을 전송 합니다. 홈 서버는 클라이언트 시스템으로부터 명령을 수신 받은 후 조명 시스템을 제어 한다. 게이트웨이 기반의 클라이언트 처리 담당인 RESTful 기술은 ‘인터넷상에 숨어있는 다수의 클라이언트들에 대한 증명’을 요청한다. 본 논문에서는 동시성 이벤트를 처리하기 위한 IoT 게이트웨이의 설계 작업을 제안한다. NodeJS의 통신프로토콜 기반의 메시지 지향 미들웨어인 XMPP는 중앙 허브를 통해 게이트웨이에 접속하여 지능형 빌딩 제어 장치의 통신 부분을 처리한다.

Integration of Big Data for Connected Cars Applications Based on Tethered Connectivity

The wireless communication technologies built-in or brought in the vehicle enable new in-car telematics services. The development of connected cars emphasizes the use of sophisticated computation framework for gathering, analyzing a large volume of data generated in all aspects of vehicle operations using Big Data technologies. Since these data are essential for many connected cars applications, the design and monitoring of MapReduce algorithms for processing vehicles data using Hadoop framework will allow to build a hosting of analytics data source. This hosting data source allows different connected cars industry ecosystem to access useful data they need to afford connected cars applications.This paper studies design steps to take in consideration when implementing MapReduce patterns to analyze vehicles data in order to produce accurate useful data that are hosted at the automakers and connect cars services providers. Experiment results show that MapReduce join algorithm is highly scalable and optimized for distributed computing than Statistical Analysis System (SAS) framework and HiveQL declarative language.

Design of Processing Model for Connected Car Data Using Big Data Technology

Recently, we have witnessed a period which things are connected to the Internet. Connected cars are currently among things connected to the Internet. Wireless communications technologies built-in or brought in connected cars enable data generated by in car sensors to be transmitted to external computers where it is analyzed. The main challenge for connected cars services providers is that the collection of same vehicle’s data such as engine temperature, engine Revolutions per minute (RPM), vehicle speed are subjected to different connected cars applications which the final purpose of each of them differs. This paper studies design steps to take in consideration when implementing Map Reduce patterns to analyze vehicle’s data in order to produce accurate useful outputs. These outputs obtained through big data technology forms a storage repository for the automakers and connect cars services providers. The proposed analytical model is based on a data-driven approach. This approach consists of collecting data sets uploaded from connected cars. Those data are then monitored based on different aspects of activity of the vehicles that we quote as “Events”. Hadoop supplements by Map-Reduce functions based reduce side joins with One-To-One joins has been deployed to process a large data and delivered useful outputs. The outputs merged with external information constitute a great insights to connected cars in order to afford connected cars applications.