A Low-Cost IoT Application for the Urban Traffic of Vehicles, Based on Wireless Sensors Using GSM Technology
- 1. A Low-cost IoT Application for the Urban Traffic of Vehicles, based on Wireless Sensors using GSM Technology
- 2. Problem Urban traffic congestion is a burning issue in many cities due to growth of vehicles in use. Which results in: • High amount of accidents • Imminent road violence • Increased air pollution • Increased noise pollution • More fuel consumption
- 3. Cont To overcome these traffic related problems an application required which – Observe Real time flow – Enhance simulation of traffic – Focus on traffic trends
- 4. Solution 1. Implemented a low-cost wireless application. 2. Supported on a distributed multilayer model. 3. 2 arduino in master-slave mode are used. 4. A laser based detection sensor used 5. GSM module used 6. Mining Techniques being applied
- 7. Functional Software Requirements • Web application, which records in the database the count of vehicles. • Web application that allows the user to observe traffic in real time. • Pentaho tool used for applying analysis on data.
- 9. Software Architecture • Software Architecture has been divided into three subsystems, having specific functions and interacting with the database. • The first subsystem is a Web application receiving data, which is developed in PHP and communicates by means of the HTTP protocol.
- 10. Software Architecture • The Second subsystem monitors that what has been developed in PHP and uses the Laravel Framework that implements three-tier architecture.
- 11. Software Architecture • The last system is a data mining process, which improves system performance monitoring by means of Pentaho. • Pentaho prepares data integration, and blends data to create a complete picture of this issue that drives actionable insights. • This platform delivers accurate, analytics-ready data to users from a monitoring device source.
- 12. Data mining and Business Intelligence • For detailed information about the behavior of the traffic at the study site, the technological module Business Intelligent platform of Pentaho has been applied (i.e. BI server) that allows to work with cubes and multidimensional databases.
- 13. Star schema to represent behavior collected by sensors. This figure indicates the Star model of the multidimensional database used. All data collected during the two week period of our study were recorded in the Mysql DBMS. Afterwards, these data were copied and transformed to the multidimensional Postgres DBMS.
- 14. Functionality Tests • Testing revealed that the functions of the software are operating, the inputs were accepted and that there has been a correct output. • Two agile methodologies as Scrum and Xtreme Programming (XP) were combined. • Scrum is responsible to the planning, implementation and documentation of the project. • While, XP focuses on the coding of software for the experiment.
- 15. Development • Development divided into three iterations • The methods used were: – Test-driven development (TDD), which is a method of software development in which unit testing is repeatedly performed on source code; – The incremental design which split a main problem into smaller problems; – The continuous integration, which main aim has been to prevent integration problems. In this last phase, several units for each of the iteration tests were carried out using Scrum. • The PHP-Unit tool of Laravel Framework has been used to generate necessary testing.
- 16. End Product
- 17. Statistical Validation • A statistical approximation has been used to the number of vehicles that pass by. • We have used the Poisson distribution, which is applicable to random events that occur over time. • Then we applied the Exponential distribution that models the elapsed time between two consecutive events modeled by the Poisson distribution, and determined its Probability and Density functions.
- 18. Results obtained by Data mining Percentage of flow by lanes.
- 19. Hourly flow illustrate per hours/ lanes.
- 20. Linear dispersion per day and per road.
- 21. Hourly flow recorded in lanes with different weather conditions.
- 22. Future Work Focus on improvement of: •Detection algorithm •Classify vehicles according to height, size & weight etc
- 23. Conclusion This system evaluate a classic problem of cities with a classically high traffic density. Benefits & Strengths are: •monitoring traffic in normal conditions •congestion traffic even in different weather conditions •Inexpensive ($200/device)
- 24. Conclusion System also contains a mining technique: •extraction of meaningful information from huge databases • allows users to make predictions that solve problems associated with urban traffic congestion. Helpful for: •researchers, •teachers and students, •Traffic Division authorities to propose strategies to redistribute the vehicular circulation