Enhancing cyclist safety: design and evaluation of early hazard notification systems for smart glasses

The number of cyclists has increased in recent years, driven by a growing preference for healthier, climate-friendly lifestyles. Despite considerable improvements in road safety in the last decades, the risk of being involved in accidents with motorized vehicles remains high for cyclists. Current strategies aimed at improving cyclist safety primarily focus on infrastructure enhancements, such as separated bike lanes, and aiding motorists in preventing collisions with cyclists, for example, through collision warnings. However, current measures often neglect the potential of directly supporting cyclists by notifying them of potentially hazardous situations with motorists.
Our main hypothesis is that early notification would allow cyclists to deliberately select and execute appropriate countermeasures after assessing hazardous situations to prevent accidents. Unlike prior approaches, our research explores an innovative way to enhance cyclists’ situational awareness of potential threats through early hazard notification. Through our research, we address (i) how early hazard notification systems for cyclists should be designed, and (ii) how using them affects cyclists’ perception and behavior.
We followed an exploratory and participatory approach, in which we increased the notification complexity iteratively to investigate different design concepts. We built on insights from simpler situations, starting with single encounters involving a single hazard, and moving toward more complex scenarios with multiple encounters and multiple hazards. To answer the above-mentioned questions, we targeted the design and effect of early hazard notification concepts for cyclists in four of our research activities that we conducted through two user studies in a bicycle simulator (N1=24, N2=32), a user study on a test track (N3=21), an online survey (N4=74) and focus group discussions (N5=7). We included simulation-based user studies, as we explored novel, not yet available technology in a safety-critical context. Consequently, in our fifth research activity, we also investigated (iii) the validity of findings about early hazard notifications obtained from simulation. We could confirm that the results from our simulation environment are comparable to those from our test track assessment.
Overall, we found that early hazard notifications provided at head-level can enhance cyclists’ safety. Our evaluations highlight both the benefits of such systems and potential challenges that need to be addressed. This thesis advances the underexplored research on direct cyclist support systems, providing a basis for future studies on early hazard notifications. Our key contributions include: (i) a set of design recommendations for early hazard notifications for cyclists, based on our research findings, (ii) experimental findings regarding the notifications influence on cyclists’ perception and behavior, and (iii) an assessment of behavioral validity regarding the effect of early hazard notifications in a bicycle simulator study.