This document describes a real-time vehicle monitoring system using Apache Spark streaming. The system aims to reduce response times during accidents by automatically detecting accident locations from continuous GPS and sensor data streams from vehicles. The data is processed using Spark streaming to identify anomalies like sudden drops in speed that could indicate accidents and monitor for speed limit violations. Detected incidents will be plotted on a map with color-coded markers to notify authorities quickly. The system was simulated by generating random vehicle sensor data and processing it with Spark streaming to analyze for possible accident locations in real-time.
2. INTRODUCTION
● Event Monitoring is the process of collecting, analyzing and signalling event
occurrences which may stem from arbitrary sources in both hardware or
software.
● Event Monitoring is used to analyse the change in the behaviour of a
particular object.
● The Object which is monitored is called the Monitored Object.
● A Monitored object must be properly conditioned with event sensors to enable
event monitoring.
3. ABOUT THE PROJECT
● Our aim is to build a Real Time Vehicle Monitoring System.
● Incidents of Road Rage have increased considerably over the last decade
resulting in many casualties.
● Most of these casualties can be attributed to the long response time required
to reach the place of accident.
● This is due to the fact that the process of determining the location of the
accident and communication with the concerned authorities is quite lengthy.
4. OBJECTIVE
● Reduce the time required to report an accident and to determine its
location more precisely.
● Monitor Real-Time GPS Data to identify possible accident locations.
● Monitor sensor data to keep a check on speed limits.
● Notifying the concerned authorities in case of anomaly or potential
accident.
5. DESCRIPTION
● The idea is to make location identification automatic in order to reduce the
response time in case of an accident.
● Vehicular Sensors are attached to each of the vehicles.
● These sensors send a continuous stream of real-time data containing relevant
information about the position(in terms of longitude and latitude) and speed of
the corresponding vehicle.
● The underlying algorithm will look for certain anomalies in data such as
sudden drop in speed of the vehicle.
7. APPROACH
● Firstly we need to simulate the vehicular sensors.
● For this we generated random data using a python script.
● This data was generated at an interval of 15 seconds.
● It contained all relevant information about a particular vehicle like its GPS
coordinates, speed etc.
● Apache spark streaming server is used to collect and analyse the incoming
data from these sensors.
8. ● The incoming data is processed by the spark streaming server.
● This data is then plotted on the map to look for possible accident spots.
● Following three kinds of markers are used in the map
○ Red Marker- This indicates the location where accident has actually
occurred.
○ Yellow Marker- This indicates about speed limit crossing.
○ Blue Marker- This indicates about the heavy traffic in a particular
location.