Road Safety

Road safety is characterized, based on the vision zero, as the absence of fatalities or severe injuries when people move on roads (including sidewalks).

Data is needed to measure safety itself, e.g., based on crashes, and on factors associated with crashes, which can be related to users, vehicles and infrastructure, as well as some elements of the larger environment such as the weather.

Safety can also be measured proactively using surrogate measures of safety.

Data Sources

Public or semi-public data sources, see the open source intelligence community.

Data for safety diagnosis often needs to be extracted from the raw data, e.g., images, through various methods including recent advances in AI (computer vision) methods.

TODO: upload sketch of data sources.

Analysis

Statistical and AI methods.