Recent studies have highlighted the significant impact of built environments (BE) in residential neighbourhoods on well-being, focusing on correlations between micro-scale BE features—such as trees, grass, fences, and bikes—and specific well-being aspects like physical health, social interaction, or perceptions of safety. However, these studies often fail to comprehensively explore how these features influence residential preferences or provide clear design guidance. To address this gap, examining residential location choice (RLC) datasets can offer valuable insights into how various micro-scale BE features affect the attractiveness of residential neighbourhoods and residents' well-being. This thesis proposes a semantic Computer Vision-enriched Discrete Choice Model (CV-DCM), which uses a panoptic segmentation model to quantify micro-scale BE features and integrates with traditional discrete choice models.

A manual evaluation of 400 masks generated by the semantic computer vision model ensured accurate quantification of BE features, enhancing choice modelling interpretation. The choice modelling results highlight the specific impacts of various micro-scale BE features on RLC, providing valuable insights for urban planners. For instance, restricting unattractive features like motorcycles in residential neighbourhoods and planting or maintaining more trees are the most attractive among vegetation. By pinpointing the most influential elements, the study facilitates the cost-effective restructuring of residential neighbourhoods to boost their attractiveness and enhance residents' well-being.

Car-following behaviour is a fundamental element for vehicle manoeuvre. The heterogeneity among driving behaviour has gained significant importance since some researchers argued that it might be relevant to capacity drop and traffic oscillations. Driver space is an area around a vehicle, and drivers will feel a rapid increase in discomfort when their comfort boundary is intruded. The response intensity of discomfort caused by spatial intrusion is represented by proximity resistance. With the development of Advanced Driver Assistance Systems (ADAS), it is crucial for drivers to feel comfortable within the implemented system. Therefore, understanding the heterogeneity of driver space has enormous potential to develop better customised ADAS. This report aims to analyse the impacts of heterogeneity on proximity resistance in car-following. The HighD dataset was processed to differentiate car-following driving styles for cars and trucks through the k-means clustering methodology. After successfully inferring the proximity resistance for each driver at each frame, the impacts of heterogeneity on proximity resistance in terms of different traffic states, vehicle types and driving styles were analysed. The results show that these three heterogeneities have different extent impacts on the proximity resistance of different drivers in car following. This study is beneficial for further investigations on the potential reasons for heterogeneous proximity resistance and the development of personalised car-following models.