1. Home
  2. RIP
  3. View Record
https://nap.nationalacademies.org/catalog/27432/critical-issues-in-transportation-for-2024-and-beyond

Assessing Cyclists' Stress on A Large-Scale: A Practical Smartphone-Based Data-Driven Approach

The primary motivation for this project is to improve street design and planning through a better understanding of cyclists’ perceived stress when they bike in city streets. The project team aims to create an “affective map”, called StressMap, by integrating cyclists’ stress levels in tempo-spatial domains. The team begins by narrowing the large field of potential stressors and measures of stress using traditional methods, such as instrumented bicycles and physiological sensors. However, such methods necessarily limit the scope of data that can be collected over time. Therefore, the project team's idea makes use of various onboard sensors of smartphones carried by cyclists to measure environmental and traffic surroundings over a larger area and time frame, enabling a large-scale stress assessment. The sensory data is then automatically collected and uploaded to the cloud server, where our deep learning model is deployed, to derive cyclists’ stress distribution over a large geographic area. This project consists of four components. First, the project team plans to measure environmental conditions when interacting with different types of infrastructure and traffic situations to quantify the perceived safety of cycling infrastructure. Using this data, the project team will identify critical environmental stressors. Second, the team will develop a smartphone app to measure those stressors, as informed by the first component, via onboard sensors. AI models will be constructed to map the sensory readings to cyclists’ stress. Third, the team will conduct a controlled-environment bike simulator study to produce a labeled dataset essential for the AI model training and validate the assessment accuracy of the proposed AI model. Finally, the team will use crowdsourcing to generate the StressMap that characterizes stressful road environments and traffic encounters for bicyclists. The derived tempo-spatial distribution of cyclists’ stress will be visualized and made publicly accessible for the stakeholders.