AI-based cyclist safety hybrid modelling for future transport network

Malik, Faheem Ahmed (2021) AI-based cyclist safety hybrid modelling for future transport network. Doctoral thesis, Northumbria University.

Text (Doctoral Thesis)
malik.faheem_phd_18020149.pdf - Submitted Version

Download (10MB) | Preview


A cyclist is a vulnerable road user whose safety is affected by several externalities. The global aim of the research is to investigate the effect of critically identified variables of rider attributes of age, gender, varied environmental condition of lighting, meteorology, and micro-infrastructure variables on the safe usage of the infrastructure for a cyclist. Presently, very few works have attempted to undertake such modelling. A novel methodological framework is developed, consisting of descriptive, statistical, artificial intelligence and mathematical approaches. Accurate prediction models are developed, and in-depth knowledge of how different variables affect cyclist safety are identified, modelled, and quantified. It is found that the variables of age, gender, varied environmental conditions, and micro-infrastructure variable are critical variables affecting the safe usage of infrastructure. These variables, both individually and in combination, impact cyclist safety. Cycling safety is a dynamic variable that varies temporally and spatially. The spatial and environmental variables have a significantly varied effect on safety depending upon the rider personal attribute. As the number of safety variables that the cyclist must conform to grows, so does the risk. The riskiest environmental conditions are exacerbated by the prevailing traffic flow regime, posing a significant safety risk to cyclists. The modelling requirement of a cyclist is significantly different from motorists. A hybrid intelligent modelling paradigm is required, as demonstrated in this research. The study results can significantly impact the route choice, modelling, and planning of infrastructure. A shift in the road safety analysis towards nanoscopic modelling can help achieve zero-vision road traffic fatality. The research reinforces a need for planning and design of infrastructure to move towards a more holistic approach while considering the limitations of this vulnerable road user.

Item Type: Thesis (Doctoral)
Uncontrolled Keywords: intelligent transportation system, mobility as a service, green transportation, intelligent mobility, smart city
Subjects: G700 Artificial Intelligence
H900 Others in Engineering
Department: Faculties > Engineering and Environment > Mechanical and Construction Engineering
University Services > Graduate School > Doctor of Philosophy
Depositing User: John Coen
Date Deposited: 07 Apr 2022 09:26
Last Modified: 07 Apr 2022 09:30

Actions (login required)

View Item View Item


Downloads per month over past year

View more statistics