Data Usage Monitoring Method Using Metadata in Vehicle Data Distribution

With rapid increase in IoT devices, large amount of data is distributed to create services that improve QoL of society. In this research, we focus on data distribution in automotive domain to increase safety, convenience, and other important benefits. However, such data distribution leads to the issue that user data can be used for inappropriate service or any other business use case. Therefore, it is important to monitor how the data is used and what services are generated after providing data. Monitoring data usage can be realized by checking the information such as data owner, data holder, service generated which is registered in metadata. The metadata is created for each vehicle data distribution. Since vehicle data is a time series data generated in large number, its distribution causes large number of corresponding metadata. Therefore, the created metadata for monitoring is large. It increases the time taken to monitor it, and hence we propose novel efficient monitoring method. This method involves creation of metadata for a data collection made over time window. Time window is selected for accident warning use case which uses Event Data Recorder (EDR) data of past accident vehicles to create warning service. To verify the proposed method, we implemented it using open source blockchain technology and evaluated by comparison with conventional approach. As a result, we found that our proposal reduced the time to monitor by 98.3% using time window of 6 seconds.