XR-Eye-Tracking for construction sites
Initial situation
What grabs the attention of people while they navigate in a space and/or perform an action and why? Answering this question has the potential to develop better training applications for novices or enable virtual robot helpers to perform tasks faster or even anticipate a person's needs. In settings where productivity and safety are critical (e.g., construction sites), one can utilize this information for deciding where machinery and tools are located, when an area is safe for people or not, and what a robot can fetch to make the process more productive or safe?
Problem statement / Project goal
What and why grabs the attention of people while performing an action/task can help increase productivity, safety, and even emotional response when performing a novel task. By analyzing the what and why for a diverse set of users, applications can be developed that utilize this information to, e.g., help new users perform a task. Your task will be to conceptualize, design, prototype and test an XR app that can facilitate such eye movement data collection and analysis.
Solution developed and its benefits
The Project provides an authentic recreation of an indoor construction site to examine how users visually examine their surroundings.
- A system that tracks and logs what users are looking at in the VR scene
- a variety of hazards users are supposed to spot
- Warning signals that aid the user in spotting hazards
- Users can mark hazards to confirm they saw and identified them
Using the Meta Quest Pro this project provides a dynamic VR simulation that allows for users' gaze data to be recorded as they are tasked with spotting potential hazards on a construction site.
To assist them the hazards can be equipped with warning signals, and by analyzing the gaze-data, the use of said signaling can be examined.
Key terms
- VR Simulated enviroment in VR
- Unity Platform the VR application is built in
- Eye-tracking Examening what users look at
- Meta Quest Pro Meta's VR HMD capable of tracking user's eye-movements
Costumer
Iro Armeni
Assistant Professor, CEE
Stanford University
https://gradientspaces.stanford.edu/
Team
Joshua Wyss
Advisors
Arzu Cöltekin