As the largest mobile telecom carrier in the world, China Mobile has the world's largest wireless mobile network, based on the existing vehicle networking equipment (CAN-bus, OBD, ADAS, equipment fatigue warning system, GPS, driving recorder, etc.), which can provide vehicle networking service, based on vehicle networking data analysis and provide users risk assessment, vehicle real-time risk monitoring, and comprehensive financial institutions for the vehicle and provide data support for differentiated financial services.
The main contents include the following:
1. Vehicle and drivers data collection: Collecting information of vehicle's mechanical status, driving behavior, and surrounding environment through OBD, ADAS, fatigue warning system, GPS, and other equipment.
2. AI technology application: mainly include the identification of the driver's body state, the wine driving, the fatigue degree, and so on.
3. To improve the accuracy and applicability of the risk assessment model through machine learning.
Business sessions: The case of vehicle networking financial services accomplished by China Mobile
With the popularity of the car in people's lives, the automobile finance service came to gain rapid development. However, the traffic accident rate began to rise, insurance fraud, potentially fraudulent incidents, how to prevent financial risks, to provide differentiated financial services of financial institutions has become the focus of concern for the different risk levels of customers. With the development of new vehicle networking technology, we can get the information of vehicle's mechanical state, driving behavior, and driver's status very conveniently and analyze the driving habits, driving status, and driving environment of users through multisource data.
This project is based on the data of car networking and mobile B/O domain data, as well as external finance, traffic police, meteorological data, through the integration of multi-source data analysis, through self-learning model to identify the risk driving behavior, issue car accident easy environment, driving fatigue, provide differentiated financial services data support, customer service and real-time risk early warning financial fraud financial services for vehicles.
The difficulty of technology is mainly based on real-time and stable data acquisition of vehicle networking, data quality problems caused by data diversity, artificial intelligence recognition, and risk assessment model of data.
Because most of the vehicles are moving and there are many high-speed mobile scenarios, data access is achieved through mobile IoT network with enhanced coverage, low latency, and high stability.
Because the factors of equipment and network lead to data missing, exception, and duplication, in order to improve the accuracy of the results, we preprocessed, filtered, and reorganized data for three times. The outlier data point monitoring, data completion, and data standardization for data preprocessing; then Kalman filtering, wavelet filtering, Gauss filtering algorithm to smooth the data; finally the signal sampling, signal interpolation /smoothing of the data reorganization, data quality assurance.
The artificial intelligence identification of the data is mainly the status recognition of the driver. Through the image recognition technology, the driver's mental state and fatigue degree are analyzed.
Because of the existing design model to assess the risk of inadequate coverage and high rate of false positives by experience, we design a machine learning model by iterative learning, K-means clustering, hybrid Gauss, convolutional neural network, deep learning, PCA, random projection algorithm to obtain the risk assessment of the parameters in the model so as to construct a more suitable risk assessment and a more accurate model.
With the help of the car networking platform for financial services provided by the insurance company claims 10,000 samples self-learning model, we provide a risk assessment report of 135 users, compared with the historical data, risk probability, and risk assessment report is consistent with the degree of 88.6%. Also through the real-time monitoring, the drunk driving customer Insurance fraud events in 1 cases.
Conclusion: the large data analysis of the car network can effectively improve the operating efficiency and reduce the financial risk of the automobile financial enterprises