Chapter 2 explores the feasibility, requirements, and challenges associated with integrating AI capabilities into AR systems to enhance the safety of highway work zones. This chapter delves into the feasibility, requirements, and challenges associated with incorporating AI capabilities into the AR system to develop a predictive safety system that can proactively identify potential hazards and issue timely warnings to workers. The outcomes of this chapter indicate that the real-time communication latency and AI execution latency meet the tight timing constraints of a real-time safety system. The early user research demonstrates positive reception and acceptance of the proposed safety framework and interface by highway maintenance and operation professionals across multiple states in the US.
Chapter 3 focuses on conducting a mixed-method usability investigation of the proposed AR-based safety system using a high-fidelity prototype. The investigation assesses aspects such as user interface design, interaction patterns, and user feedback to evaluate the overall usability and effectiveness of the technology in enhancing roadway work zone safety. The findings indicate that participants rated the usability of the system above average in both indoor and outdoor settings and perceived a reasonable level of mental effort. Perceived trust was found to be significantly correlated with usability, underscoring its importance in user experience.
Chapter 4 examines the impact of different sensory modalities on worker reaction times in augmented reality warnings within roadway work zones. The analysis of data from experiments provides insights into the effectiveness of various warning modalities, including visual, audiovisual, haptic visual, and combined haptic audiovisual cues, in improving worker response times. The findings indicate that the haptic visual design elicited the fastest response on average among the participants, and its performance was comparable to that of the audio haptic visual design. Furthermore, both of these designs demonstrated significantly faster reaction times compared to visual and audiovisual warnings. The results also indicate that reaction times to augmented reality warnings in real-world outdoor scenarios were generally longer and exhibited greater variability compared to baseline desktop warnings and simulated AR in virtual reality. Surprisingly, VR simulated warnings did not show statistically significant shorter reaction times compared to their real-world counterparts. These observations suggest that simulating AR in virtual reality may not accurately replicate the reaction times observed in real-world situations.
Collectively, the results from these chapters demonstrate the usability, perceived safety benefits, and potential for timely notifications offered by the proposed AR-based safety system. This research also contributes to establishing best practices for designing time-sensitive safety systems, prioritizing situational awareness, and implementing worker-centered design principles in AR safety systems. Ultimately, the findings have the potential to significantly enhance the safety of highway workers and the broader workforce operating in roadway work zones.
-
Prospective Students
-
Faculty and Staff
-
Current Students
-
Alumni and Friends