BUS TRACKING AND MONITORING SYSTEM BY USING IoT

Transcription

1 BUS TRACKING AND MONITORING SYSTEM BY USING IoT Dheepika.D #1, Gunasundari.P #2, Rubitha.B #3 Santhoshkumar.M #4 1,2,3 UG scholar, Dept.of EEE, 4 Asst. Prof. Dept. Of EEE 1-dheenanmadheswari@gmail.com, 2-pammuguna28@gmail.com, 3-rubithabalashanmugam@gmail.com 4-shannewat6@gmail.com AVS College Of Technology, Salem, Tamil Nadu, India. ABSTRACT In this paper the Tracking of transport system is done for each station and its route distance is done using NodeMCU microcontroller. Tracking System involves the installation of an electronic tool in a bus, with an installed Android App on any SMART phone to permit the Administrator/User to track the bus location. There are two applications one for server and the other for the client. Buses carry GPS devices to track their positions. By this positions to server are periodically updated. Client application displays map showing the position of bus. It shows where buses are on a map and provide students and staffs the updated information at different time interval. The server will monitor location and will store its data in the database. It is a real-time system as this method automatically sends the information on the GPS system to a system/smart phone. And it is connected to the cloud and following through the android app. The waiting time of the user can be reduced. Simple mode of communication is the key feature of the Bus Tracking system. This application can be easily extended for central tracking system to keep track of all the buses. Keywords Vehicle tracking; microcontroller; RFID; GPS;GSM/GPRS; Android; Cloud. I. INTRODUCTION Ensuring safety and security of student is a prime concern for society, the world over. A recent study delaying of bus shows that number of missing students, across the country, increased by 84 percent between 2013 and Technology can provide a comprehensive solution to this vexing problem. Bus tracking systems play a major role in the safety of student. For tracking a bus we propose an IoT-based bus monitoring system. IoT is defined as the inter-communication of "connected devices" and "smart devices", and other items embedded with electronics, software, sensors, actuators, and network connectivity, which enable these objects to collect and exchange data. [2]. An in-vehicle device, using Radio Frequency Identification (RFID) tag, identifies each student and tracks the bus via the global positioning satellite (GPS) module. A GSM module is deployed to notify the parents about the entry and exit status of students, and up to the minute reports, of running status of the bus, appended with location. Short message alerts are sent to parents, if the predefined, route schedule of the bus, is modified. The information collected by a GPS receiver in the vehicle is communicated to a Cloud server. A smartphone application empowers the end user, to monitor the location of bus, and view the route path, using Google maps. The mobile application also determines the arrival times of a Bus, using a Kalman filtering-based prediction algorithm. The mobile app has the ability to continuously update the transit times, coordinated with location identification of GPS modules. It uses historical data and real-time information, for computation of the arrival times [3]. Archived bus transit history can be useful to both parents and Authorities, in future analysis, towards ensuring the safety of student. This paper provides a framework for integration of both hardware (tracking device) and software subsystems (Kalman filtering algorithm and Android smartphone application). The proposed work focuses on the design and performance analysis of the bus tracking system. The collected data is stored in IoT Cloud platform for processing at a later date. The rest of this paper is organized as follows: Section II discusses related work in this domain. In Section III, we give a detailed description of the proposed system. Section IV presents description of major components of the tracking system. Section V describes the methods for predictions of bus arrival times. Test results and analysis is described in Section VI. Closing remarks are noted in section VII. II. RELATED WORK This phase of the paper elaborates telematicstechnologies for the progress of various vehicle monitoring applications like tracking location of vehicle. This will allow parents and authority to keep real-time track of the bus behavior. A publish and subscribe architecture [4] is used; a parent or any stakeholder, who needs data is called subscriber. S. Lee, G. Tewolde and J. Kwon developed and tested a vehicle tracking system, in this paper an invehicle device comprised of GPS, GSM and General Packet Radio service (GPRS) modules, and it is embedded inside the vehicle, whose position is to be tracked in real time using a smartphone application. A hardware prototypewas developed for vehicle tracking in[5]. Geographical coordinates of the vehicle, obtained by GPS Receiver, are communicated to the user s phone, using GPS-based Bus Tracking System. It is an application that has its client side on the Android platform and the main objective of this paper is to acquire real-time location of the bus and the updated transit schedule of the bus-route, it is comprised of two main modules; a GPS-based sub-system, which tracks the current location. To make

2 efficient tracking system some papers have introduced features like bus arrival time calculation. Liu, et al in[7]integrated the k-nearest neighbor (k-nn) algorithm with cluster analysis and applied principal component analysis (PCA) for computations of bus arrival times. Road traffic is the key factor that effects arrival time calculations. In [8], Sai, et al, dealt with solutions to different traffic anomalies using Hadoop map Reduce. For more precise prediction of the road traffic a cost effective method using Apache Spark was introduced by Prathilothamai, et al in[9]. III. PROPOSED SYSTEM The proposed system consists of both hardware and software module. The hardware layer is composed of a tracking module. The core part of the tracking module is Arduino microcontroller with various peripherals including RFID reader, GPS receiver and GPRS/GSM modem. On receipt of the data at microcontroller, it is processed and transmitted to Cloud storage for future reference. The software module is comprised of a arrival time prediction and a client side application. Client side application allows the user to view the dynamic movement of the vehicle with predicted arrival times. Client side application was developed on an Android platform. Fig.1 shows the threelayer architecture diagram of the system. it to microcontroller for further processing. The tracking device is placed near the door of bus so that an alert message will be sent to parents whenever a student gets into the bus or leaves the bus. GPS is used for tracking process. To estimate the accurate arrival time of the bus, a Kalman filtering dynamic algorithm is used. A. Hardware Layer The system hardware an Arduino Mega microcontroller (MCU), GPS receiver, RFID reader module and a GPRS/GSM Modem. The RFID reader module is used to read the RFID tag data. Here we use a passive reader. Once the data is received by the reader via the serial port of Arduino, The microcontroller will extract the relevant data from GPRMC string. After processing the data from GPS and RFID reader, microcontroller will transmit the data to cloud storage through GPRS modem. The various hardware components of the deviceare explained below. 1)Microcontroller: The Arduino mega microcontroller serves as the core part of the tracking module to control the bus tracking system. A C language program, saved in the microcontroller s memory, controls the modules,. The Arduino Mega 2560 has 54 digital input/output pins, 16 analog inputs, 4 UARTs. The proposed system uses 3 UARTs-one for RFID, one for GPS Receiver, and the third for GPRS/GSM Modem. RFID Application Layer Technologies GPS GPRS Communication layer 2) RFID Reader: One of the key hardware components in the system is EM 18 RFID Reader. This is the most commonly used RFID reader to read 125 KHz tag. It contains an antenna that can be powered by 5V power supply. RFID) is an automatic identification technology where digital data is encoded in an RFID tag. The reader is a radio frequency (RF) transmitter and receiver, controlled by a microprocessor/mcu. The antenna attached inside the reader captures data from tags, then passes the data for processing. Tags are issued to each student, together with their roll numbers. When the tag is placed near the reader, it will get energized and data is transferred to reader, using radio waves. SIM28MLGPS RFID Reader Hardware layer Microcontroller GSM/GPRS modem Fig.1: Architecture diagram. A RFID System exists with a RFID tag along with a unique Id. RFID card reader will read this unique Id and give 3) GPS Receiver: The proposed system uses SIM 28 ML GPS receiver which has excellent low power consumption characteristics. The module has complete signal processing from antenna. This module requires 12V power supply. The host port is configurable to UART. Transmitter pin of modem is connected to receiver pin of Microcontroller. The Global Positioning System (GPS) helps to provide geographic coordinates of object anywhere on Earth with exact Universal Time Coordinated time (UTC). GPS offers great accuracy and generate real-time tracking location data. GPS satellite transmits the location data. GPRS receiver receives the signals innmea format. There are different NMEA signals. 72

3 We are focused on GPRMC signal, since it contains location information along with the speed. Form of GPRMC is given below: $GPRMC, , A, , N, , E,022.4,084.4,230394,003.1, W*6A 4) GSM/GPRS Modem: SIM900 is designed with a very powerful single-chip processor. Transmitter and receiver pins of modem are connected to receiver and transmitter pins of microcontroller. Modem requires 12 V power supply. The GPRS module is responsible of establishing connections between an in-vehicle device and a cloud for transmitting location information of the vehicle [10]. B. Communication Layer Communication layer is used to establish a connection with underlying hardware, the MCU and the upper layer applications. Various underlying technologies used to develop the application are RFID, GPS and GPRS. Various technologies used for the communication are explained below. The proposed architecture uses the following algorithm at the hardware level. Transmission and Reception Algorithm at the Microcontroller Input: UART 1 Rx pin (RFID Reader OUT), UART 3 Rx Pin (GPS Receive OUT) Output: UART 2 TX Pin (GSM IN) 1. Read the 12-character RFID data from EM 18 using Microcontroller. 2. Read NMEA data from the GPS. 3. Extract Date, Time, Latitude, Longitude and speed from NMEA data. 4. Send the message to the recipient/cloud server using AT command through GSM modem. C. Application Layer 1) Mobile Application: The latitude and longitude coordinates are transferred to the central server. A mobile application has been developed to access location data and display the vehicles movement in Google map, using Google map API. Users can retrieve information, where users select the bus route number and date, and receive the transit time schedule. UBIDOTS is a cloud service to store and analyzes information from sensors in real time. It enables creation of applications using IoT. While sending sensor data to Ubidots, it can easily create widgets for viewing in real time, send short message sequence (SMS) or Notification. In this project, RFID tag values along with GPS data are sent to Ubidots. The software module is comprised of arrival time prediction and a client-side application. Client-side application allows the user to view the dynamic movement of the vehicle with predicted arrival times. Client-side application was developed on an Android platform IV. DATA COLLECTION AND ANALYSIS In the present study, data was collected using GPS units fixed within buses. The route selected for the study was Salem to Ayodhiyapatinam, which spans 25Km. There are 9 bus stops on this route. Every 45 seconds, GPS data, from the device, installed in the buses, was sent to a server. GPS data included the latitude, longitude, speed and time stamp. Data exchanges took place in real time, through GPRS. This data packet, when received by the server, was conveyed to Ubidots cloud as well as appended to a.csv (comma separated value) file. The collected data is shown in TABLE I. The data received in the server was processed immediately, in order to make a real-time prediction of the arrival time of the bus. The data from previous days, and the current bus-stop s real-time data received from GPS receiver, were used to predict the arrival time of next busstop. The distance travelled, between two consecutive time intervals, was calculated using Haversine formulae, which computes the accurate distance between two places using latitudes and longitudes. TABLE I. HISTORICAL DATASET COLLECTED BY TESTVEHICLE READING ON 02/03/2017 MORNING Place Latitude Longitud Time Speed Salem eeee am 30km/hr Nagyarkulanga am 32 km/hr ra Kanjoor am 31 km/hr Cheppad am 32 km/hr Ramapuram am 35 km/hr Kalarikkal am 30 km/hr Kareelakulanga :01a 32 km/hr ra Edasserry jn m 8:17a 34 km/hr Valapadi m 8.22am 33 km/hr V. KALMAN FILTERING 2) Cloud Server: 73

4 In the proposed framework, we used Kalman filtering algorithm to build the prediction model. Kalman filtering is a powerful method for predictions, in real time systems, since it estimates the state of a system from measured values, which include errors[11]. This is an iterative algorithm, which involves two stages that first predicts the system, and then, corrects the same using measurements. Kalman filtering has two steps: 1. Prediction 2. Measurement update. During literature review studies the basic algorithm for the Object Tracking on the Global Positioning System; this entire algorithm is described as bellow: Assume that the prediction Localization Improvement Algorithm: This localization improvement algorithm in GPS interfering spots by integrating information of multiple sensors such as gyroscope and compass in smart phones. The proposed algorithm is implemented in a smart phone [12]. There are 3 steps to implement this algorithm: 1. Structure of Algorithm:- The proposed algorithm consists of three steps: 1. Finding the correct direction of movement. 2. Obtaining the distance moved, and 3. Integrating the prior results and GPS location information. To obtain the direction of movement, get the heading from the built-in compass in the smart phone. However, the compass is highly dependent on the ambient magnetic field so it has lower accuracy. To obtain more accurate values of the user's heading, we stabilize the heading value of the compass by recurrence processing of the data. After obtaining the direction of the movement of the user, the distance moved is calculated by the summation of the distances of each coordinate from the GPS location information. 2. Recurrence Processing Function of Compass Heading x0=a0 x1= (x0 + a1)/2 xn =(xn-1 + an )/(n + 1) xi: accumulated heading value ai : a new compass value The measurements of the compass in Smartphone are highly dependent on the ambient magnetic field so its accuracy is low. For this reason, here stabilize the measurements of the compass by averaging. x0 =a0 x1 = (x0 + d1 + a1)/2 xn =(xn-1 + dn + an )/(n + 1 ) xi: accumulated heading value ai : a new compass value di : difference of heading The heading value of the compass can be stabilized. But the real heading of a user varies continuously as the user moves. To adjust this value, here apply the difference between the previous heading and current heading and process it recursively. Kalman Filter Algorithms Kalman filtering, also known as linear quadratic estimation (LQE).Kalman filter is implemented to reduce GPS errors and thus it increases the accuracy of the localization system. The algorithm works in two-step process. 1. In the prediction step, the Kalman filter produces estimates of the current state variables, along with their uncertainties. 2. Once the outcome of the next measurement (necessarily corrupted with some amount of error, including random noise) is observed, these estimates are updated using a weighted average, with more weight being given to estimates with higher certainty. 3. The algorithm's recursive nature, it can run in real time using only the present input measurements and the previously calculated state and its uncertainty matrix; no additional past information is required. VI. RESULT AND ANALYSIS In the proposed framework, a android application was provided to each parent and the application is to installed in parents mobile phone. When the parent enters Bus route number and Date, the Bus tracker shows the Bus s current location along with expected arrival times at each bus 74

5 station. Such information is very handy for parents, who need not anxiously wait for uncertain durations, as they would have acquired the most up to date arrival time of the specified bus. The following section shows screenshot of two modules. Whenever the parent clicks anywhere on the map, the current position and the travelling path of bus will be displayed, as shown in Fig.3. Fig.4: Scheduled and predicted arrival time of bus. VII. CONCLUSION We developed a bus tracking device, in collaboration with RFID, GPS and GSM/GPRS technologies. Whenever a student enters/exits the bus, the tracking system is capable of notifying the student s parents, through SMS alerts. An Android application has been developed to display vehicle location on Google maps and, display arrival times of a bus, at each of its designated bus-stop. The implementation is very cost-effective, as it is based on easily accessible electronic devices. To predict the expected arrival time of bus, at each node (stop), we resorted to Kalman filtering dynamic algorithm. The accuracy of the algorithm calculated using RMSE was shown to be self-evident as the algorithm s predictions returned the most accurate results. We found that a single algorithm cannot predict 100% accurate results. We assert that the accuracy of the proposed system can be enhanced by unifying the algorithm with other machine learning algorithms, such as Artificial Neural Network (ANN), Support Vector Machine (SVM). VIII. REFERENCE [1] S. Shah and B. Singh, RFID Based School Bus Tracking and Security System, pp , [2] J. Zambada, R. Quintero, R. Isijara, R. Galeana, and L. Santillan, An IoT based scholar bus monitoring system, 2015 IEEE 1st Int. Smart Cities Conf. ISC2 2015, [3] A. F. Abidin and M. Kolberg, Towards Improved Vehicle Arrival Time Prediction in Public Transportation: Integrating SUMO and Kalman Filter Models, Proc. - UKSim-AMSS 17th Int. Conf. Comput. Model. Simulation, UKSim 2015, pp , [4] S. Lee, G. Tewolde, and J. Kwon, Design and implementation of vehicle tracking system using GPS/GSM/GPRS technology and smartphone application, 2014 IEEE World Forum Internet Things, pp , [5] L. Singla, GPS Based Bus Tracking System, [7] T. Liu, J. Ma, W. Guan, Y. Song, and H. Niu, Bus Arrival Time Prediction Based on the k-nearest Neighbor Method, 2012 Fifth Int. Jt. Conf. Comput. Sci. Optim., pp , [6] T. Bojan, U. Kumar, and V. Bojan, Designing vehicle tracking system - An open source approach, 2014 IEEE Int. Conf. Veh. Electron. Safety, ICVES 2014, pp , [7] A. Sai, S. Salim, P. K. Binu, and R. C. Jisha, A Hadoop Based Architecture Using Recursive Expectation Maximization Algorithm for Effective and Foolproof Traffic Anomaly Detection and Reporting, vol. 10, no. 55, pp , [8] M. Prathilothamai, A. M. Sree Lakshmi, and D. Viswanthan, Cost effective road traffic prediction model using Apache spark, Indian J. Sci. Technol., vol. 9, no. 17, [9] H. Yaling, The Design of Monitoring System Based on GPRS, 2016 Int. Conf. Robot. Intell. Syst., pp , [10] N. R. Nair, 2-D Airborne Vehicle Tracking Using Kalman Filter,