Forestry Workforce Location- and Wearable-based Activity Recognition to Quantify on-the Job Digital Health and Safety Metrics

Logging is among the most dangerous professions in the United States. Manual felling of timber with chainsaws and setting of cable log chokers accounted for 47% of injuries in Idaho between 2011-2014. Building on a recent PNASH pilot project, a library of wearable- and location-based human activity recognition (HAR) models will be developed and coded into a smartwatch app prototype to enhance the safety and efficiency of forestry work in Idaho, Oregon, and Washington through increased situational awareness (SA) among workers on remote cable logging operations.  

This project will: 

  1. develop HAR models needed to classify logging work activities in real-time using smartwatches;  
  2. code and evaluate those models using a smartwatch app prototype; and  
  3. evaluate the smartwatch app jointly with a location-sharing system for active forest operations to provide both location- and smartwatch HAR-based information for improved SA to reduce logging injuries.  

HAR models will be developed and evaluated for three core job tasks on cable logging operations: manual felling of timber, manual setting of chokers during yarding of trees, and heavy equipment operation. Project deliverables include a prototype smartwatch HAR app, four peer-reviewed publications, an outreach publication, training workshops for loggers and forest managers, and a dedicated social media account. We expect outcomes of this project to include a reduction in the risk of work-related illnesses and traumatic injuries among forestry workers. 

Progress to Date

The first year of this project has primarily been preparing for multiple field studies taking place in fall 2023 and early spring/summer 2024. These field studies include time studies and collection of smartwatch, smartphone, and real-time worker and equipment GNSS location data needed to conduct the initial analysis for the project. In Year 1, we purchased the primary equipment necessary to complete this proposed work, including smartphones and smartwatches. We have upgraded and maintained a variety of existing electronic devices that are also needed, including digital radios that create the network used to improve communication and safety among loggers at the jobsite. An outstanding Postdoc, Dr. Eloise Zimbelman, was hired for the funded postdoc position on our project. Dr. Zimbelman wrote computer code to update and improve on an existing app that will
be used to collect the Garmin smartwatch data needed for the research. As of August 2023, the new code is working well and we are now able to begin our primary field sampling for the project.

Next Steps

In the next year, we will be conducting observational sampling on active logging operations in order to collect the primary field measurements needed to create models of work activities and to begin coding the new smartwatch app to improve the safety of
loggers. We will be sampling at 5 different field sites in Idaho, Oregon and Washington. The data collected will be processed, analyzed, and used to begin preliminary app programming. Additionally, we will work with our project Technical Advisory Board to
gain input on our methods, on the best potential uses of the new app, and on discussion of ethical considerations. Lastly, we will begin a longitudinal survey intended to characterize the perspectives of logging contractors and workers on the use of smartwatch apps for logging safety.

Partners and Collaborators

Principal Investigators: Robert Keefe, Associate Professor and Director, University of Idaho Experimental Forest, and, Eloise Zimbelman, Postdoctoral Scholar, PNASH Center and University of Idaho 

NIOSH Funding Period 2022-2027