This demonstrator explores how vehicle driving data and other road safety related data can be meshed and modelled, aggregated and semantically annotated in order to extract meaningful, automotive and road safety-relevant information for various stakeholders.
Scenario
The automotive and road safety demonstrator will be developed according to three different scenarios, Broken Road Indicator, Safe Driving Indicator, and Regional Driving Style Risk Estimator. The three different corresponding versions of the automotive and road safety demonstrator are then aimed to provide the following benefits to the users of the services:
- Provide insights into road conditions based on exploiting individual vehicle sensor data, traffic data, and map data (Broken Road Indicator).
- Infer the driver’s safety style and then calculate a safety index, through utilising vehicle sensor data along with environmental information and other content (Safe Driving Indicator).
- Calculate a regional driving safety risk metric for certain regions including intersections, streets, cities or countries (Regional Driving Style Risk Estimator).
The final automotive and road safety demonstrator will include all three versions, Broken Road Indicator, Safe Driving Indicator, and Regional Driving Style Risk Estimator.
Data Sources
The system will analyse and process domain-specific data coming from various sources, including, yet not limited to, (anonymised) driving data, map data, and other proprietary or external open sources (e.g. weather data, social media data, map data, or traffic data).
Architecture and Operations Summary
The demonstrator architecture is graphically illustrated in the following schematic. As per the illustration, the core module responsible for the aggregation of the data from various sources is the AEGIS platform, which feds data from open sources as well as from platform end users.
Capabilities for conducing data science and data analysis are continuously demanded by the mobility domain. Hence the AEGIS platform plays a fundamental role as it provides the required data analysis infrastructure for the data scientist at Virtual Vehicle (VIF).
The automotive demonstrator will be located in Greater Graz area in Austria as the majority of vehicle trips will be recorded in this area. The VIF PSPS data scientists will use the AEGIS platform as basis for their different analyses. Data from various sources will be processed by AEGIS platform enabling services for various stakeholders including drivers, city planners, road maintenance departments, or automotive engineers. Services to be developed in the AEGIS project include identification of road damage (from vehicle data), identification of safe driving events and calculation of an individual driving risk score (from vehicle data and weather data), as well as estimating an aggregated regional driving risk (from safe driving data, risk score data, vehicle data, and weather data).
The visualization above shows project-relevant data sources, direct users of the AEGIS platform (PSPS data scientists at VIF), as well as beneficiaries of the provided services (drivers and city planners) in a simplified way. Analysis is provided as a service by the VIF data scientists to various service consumers (drivers and smart city planners to name two concrete stakeholders) ‘offline’ in a first step. While drivers in return provide (their) driving data in an anonymized way, city planners will have to pay a fee for service consumption. An online user-self-service functionality is envisaged as a future exploitation.