Multimodal Data for Road User Behavior Analysis to Support Safe Driving Patterns

Project manager
Yacine Atif
2021-01-01 - 2021-09-30

Human-operated vehicles with Advanced Driver-Assistance Systems (or ADAS) are projected to
prevent human-failures at the wheels. To support this purpose, driver state is estimated using
both internal factors such as fatigue, inattention and drowsiness levels, as well as external
roadway factors such as traffic volume and roadway configuration. The objective is to
formulate a predicted safety score based on the data at hands, to support automotive safety
solutions for car manufacturers and OEM providers, as well road safety at large, for
infrastructure authorities.
New synthetic information resulting from fusing authentic in-cabin and infrastructure data
contribute to a holistic safety score. The project builds a conceptual framework of risk
prediction involving driver behaviours under different road condition settings. This framework
could also augment mobility simulators to assess traffic under realistic safety settings.

Short facts

Project title: Multimodal Data for Road User Behavior Analysis to Support Safe Driving Patterns
Project type:
Research area:
Road user behaviour

Project publications

Recognizing driver behaviors dynamically in safety modeling is a challenging problem since it involves various feature parameters about the driver, the car and the ambient traffic. The goal is to cateogrize driver behaviours into a standard
Yacine Atif
Year of publication

Safer – Vehicle and Traffic Safety Centre

SAFER is the open research arena where researchers and expertise work together to create safe mobility. Our traffic safety approach covers people, vehicles and the infrastructure – and together we contribute to safer road transports and smarter, more sustainable cities.

Contact information


Lindholmspiren 3A
SE-417 56 Göteborg

 +46 31-772 21 06
 safer [at]