21-10-2016, 09:23 AM
1460232935-section3.docx (Size: 20.95 KB / Downloads: 5)
INTRODUCTION
Nowadays, all the smart phones have inbuilt sensors and processing power. The essence of thisproject is to collect real time weather data from using these sensors. The data is collected from smart phones and processed according to the user needs. This data can be very useful in forecasting and analyzing weather conditions, humidity levels etc. All these devices are connected to a cloud environment and useful information is obtained from it. In this project, a mobile sensor cloud infrastructure is implemented. The user is presented with a dashboard on the client side which displays data processed sensor data. The cloud interface is used for the processing of data. The data received from mobile sensors is stored in the cloud database. Each Node helps in receiving data from the mobile sensor devices. The Data Management is also performed in the client side. The project can be divided into the following components:
1)Cloud infrastructure management:
This service acts as an interface between users and mobile sensors.The data processing is done in this component.
2) Dashboard:
User can add, delete and modify the sensor data using this web based user interface. User can visualize the collected data in required format.
3) Sensor data hub:
The collection and storage of data from the mobile sensors is done in this component.
Openstack is used as a platform in order to maintain the cloud infrastructure. It also acts as an interface between the sensor data collection, management and displaying on the User interface. The mobile device acts as a thin client and the cloud acts as the processing layer. The entire processing of data takes place in the cloud and the mobile acts as a real time sensor that provides data to the cloud. In this project, we are using mobile phones as the mode of communication. Every sensor is attached with a Mobile card and the IMEI number acts as the unique Identification number for the sensor.
INTRODUCTION TO LITERATURE REVIEW
In his paper, Wang et al. [1] estimates that 13.3% of crashes were due to what was considered distraction. 9.7% of crashes belonged to the category which said, “looked but did not see”. If drowsiness of the driver is also taken into account, this number increased to 12.3%. In [2], driver distraction was the root cause for 80% of the crashes, 65% of which were all near crashes. Evidence have suggested that driver distraction and inattention are factors contributing majorly to vehicle crashes and incidents. The National Highway Traffic Safety Administration (NHTSA) has estimated that, some form of inattention is involved in 25% of the crashes. Based on studies from NHTSA in the US and some European studies show that driver distraction has resulted in as many as 5000 fatalities and damages amounting to $40billion every year [3].
2.2 DEFINITION OF DISTRACTION
The possibility to reliably recognize and detect the state of the driver and provide him assistance is the challenge here. For instance, a forward-collision assistance system can be triggered based on the driver’s state. Based on the type of distraction detected, the strategies of the function are adjusted. Smart assistance means, to recognize the driver’s intension and state and providing suggestions without interrupting with false alarms or interventions. A number of recent methods were aimed at estimating the distraction of driver and modeling fundamentally based on stress.
There are several definitions of distraction, each having a different perspective. Most of the times, they are overlapped with the definition of inattention or other states of driver such as workload, stress, drowsiness, etc. In this paper, we stick to the most relevant definition, “Driver distraction is the diversion of attention away from activities critical for safe driving toward a competing activity.”