Increasing Work Zone Safety: Worker Behavioral Analysis with Integration of Wearable Sensors and Virtual Reality
Overview
According to the Federal Highway Administration (FHWA), work zone fatalities at road construction projects account for up to 3% of all workplace fatalities in a given year [1], and the primary causes are runovers/backovers, collisions, and caught in-between mobile equipment. One of the main proactive approaches adopted by construction companies to prevent these incidents is safety training courses, which are designed to help increase workers’ awareness of hazards around the job sites and take timely actions to avoid injuries. However, work zone safety knowledge from training courses is not enough to change the level of vigilance of workers, which is easily affected by factors such as fatigue or environmental distractions. With the development of wearable technologies, an increasing number of research studies have been exploring the feasibility of using wearable sensors to detect workers’ attention and vigilance towards job site hazards. However, merely measuring workers’ awareness of hazards is not sufficient. There is still a need to understand key parameters that impact worker and driver behaviors regarding received alarms/warnings/notifications and design notification systems that are calibrated for the optimal time, frequency, and modality to push information on potential hazards at work zones.
With the goal of reducing the number of injuries and fatalities, this project aims to understand the key parameters (e.g., work zone location characteristics, personal vigilance levels, types of construction work) that play roles in achieving responsive behaviors in workers. Key questions this research will address include in what conditions people ignore or respond to warnings, how notification systems can be calibrated for getting responsive actions from workers, and what modalities, frequencies, and timings of pushing notifications are most effective. Through wearable sensors and realistic representations of work zones in virtual reality, we plan to collect worker behavioral and physiological (heart rate) responses to warnings issued under various realistic scenarios and various warning mechanisms.
Objectives
This project’s objectives include:
- Define the key factors that influence the reaction of workers and drivers in and around work zones to warning notifications received
- Evaluate the effectiveness of wearable sensors and virtual reality in determining the key factors that influence the reactions of workers and drivers to notifications
- Calibrate notification mechanisms using reinforcement learning
To achieve these objectives, the research methodology includes:
- Development of a mobile application to be deployed on smart watches/wearable wristbands to collect integrated data on driver/worker choices and bodily states (heart rate, body temperature)
- Design of virtual environments simulating highway work zone scenarios by changing work zone location characteristics and type of construction activities
- Development of a reinforcement learning model that builds off of the collected data on simulated scenarios to minimize incidents
A comparison on real-world work zone (left) and the created VR environment (right)
Work zone point cloud before (left) and after (right) registration.
Measurements of work zone
Overview of the proposed research for increasing work zone safety for construction workers
Deliverables
- A set of key factors that determine human responsive behaviors in highway work zones.
- The evaluation of whether data from wearables are meaningful to correlate with human responses to notifications
- A guidance to effectively integrate wearables to work practices at job sites and define a sequence of steps to follow to analyze the collected data.
- A mobile application that can collect worker/driver choices (dismiss or react to a notification) along with their bodily responses (heart rate)
- A reinforcement learning model that builds off of the collected data on simulated scenarios with an objective to minimize incidents.
Application screenshots for: splash screen(left), simulation screen(middle), and alert screen(right)
Traffic simulation in bird-eye view (left), and in construction worker view (right)
C2SMART Research as part of the “Increasing Work Zone Safety: Worker Behavioral Analysis with Integration of Wearable Sensors and Virtual Reality” project was presented at the US Department of Transportation exhibition booth in the TRB 2020 convention’s exhibit hall. Ph.D. Candidate Zhengbo Zhu of NYU led the demonstration, which uses virtual reality to simulate the experience of construction/maintenance workers on active roadways. In this project, the feasibility of notifying workers for imminent dangerous events through a wearable device is being investigated, and the VR setup allows for a better understanding and testing of this technology without entering active work zones.
Personnel
Principal Investigator | Dr. Semiha Ergan, NYU |
Funding Source | C2SMART Center and NYU |
Total Project Cost | $108,148 ($60,000 from C2SMART; $48,148 matching from NYU) |
USDOT Award # | 69A3551747124 |
Start and End Dates | 03/01/2019-05/31/2020 |
Implementation of Research Outcomes | The outcome of this research will help to calibrate when, at what frequency, and how to (with what modalities) share warnings with workers involved in active work zones for effective responses towards reduction of incidents. |
Impacts/Benefits of Implementation | The proposed research can potentially help mitigate highway work zone incidents (i.e., injuries, fatalities, and near misses) by determining when, how and at what frequency to push alarms/warnings/notifications to habitants of work zones. The proposed VR based hazard simulation can serve as a safe way to capture worker response under hazardous conditions without putting them in danger. |