CLOUDIA - Methods for efficient searching of events in large volumes of data for analysis and development of ADAS

Purpose and goal:

The purpose is to develop scalable methods and techniques for defining specific traffic events and to make these searchable in large amounts of high-resolution data that has been collected and uploaded from a large number of vehicles.

The project´s objective is to (1) develop methods and techniques for defining complex events, (2) implement a research platform in five test vehicles, (3) integrate a cloud service for advanced search for interesting events, and (4) demonstrate the search function for analysis of specific events focusing on preparatory behavior and the use of ADAS.

Expected results and effects:

The project is expected to result in methods that make product development more efficient. Searching for specific events in large amounts of data can be useful for vehicle manufacturers and system suppliers, as well as for insurance companies and also for, for example, traffic and community planning (eg creating “digital twins”). The method for defining events will be made publicly available so that other actors can apply the same methodology / definitions. The new search features will also allow developers to "subscribe" to specific events in real time.

Approach and implementation:

Methodology for defining specific events is developed and made available to other actors. Vehicle platform for data collection used in previous project upgraded so the system becomes more scalable. The vehicle platform is installed in five test vehicles.

Simulations are made to see that the system is able to scale up. A function for searching for specific events in large amounts of data to be implemented in the cloud solution used for data collection. The research platform is demonstrated using the searches for specific events for analysis of the use of ADAS.