DC3. AI for road safety in LMICs

Hosts: FPZ & AGILYSIS

Objectives:

• Define a framework for exploiting available data sources and adapting existing road survey databases (e.g., SLAIN) inLMICs
• To perform AI-supported knowledge discovery in databases utilising extraction of implicit, previously unknown, and potentially useful
• information from data collected for other purposes
• Facilitate transport modelling with AI deep learning to establish predictive crash models to support transport infrastructure investment
• decisions in LMIC on road networks where traditional data is non-existing or unreliable

Expected results:

• Identification and classification of potential structured and unstructured big data sources for the development of road safety
• performance assessments and predictive crash modelling
• Innovative and cost-effective data acquisition, processing and analysis methods to support road safety management in
• LMIC (performance indicators, star rating, risk mapping, road safety audit and road safety inspection)
• A taxonomy of data, methods and tools tailored for AI road safety support to less resourced countries

Planned secondment(s):

• EIRA-SI, Supervisor: Purpose: to test the developed methods on datasets from LMICs