Investigating Tools and Methods for Data Capture of Forestry Workers

The forestry industry in New Zealand has been in the spotlight for many years due to its poor health and safety record. Workplace locations are often remote requiring extended travel times and worksites rarely offer any shelter from the elements. The nature of the work is physically demanding, requiring lots of ‘boots on the ground’ time, as such employees can be susceptible to impaired performance due to workplace fatigue. In order to assess the impact of activity on performance we initially investigated the relationship between activity measured as steps and simple reaction time. Using Fitbit Charge HR activity trackers we collected physiological data and tested reaction time at various times throughout the day. We identified that an inverse curve relationship existed between activity and reaction time; slower times being recorded at the start and end of the day with faster times through the middle of the day. We then expanded the scope of our study to incorporate individuals working in the forestry industry. PF Olsen a forest management company, sourced sub-contractor crews currently undertaking tree felling and harvesting operations within the Bay of Plenty. Participating crews used separate operational techniques; mechanised by way of plant, manual using a more hands on approach, and hauling, a combination of the two aforementioned techniques. Using the same activity monitors as in our preliminary study we collected physiological data off crew members as they performed their daily duties. In order to assess the possible impact of fatigue we measured both simple and choice reaction time at three periods during the working day; start of shift, after four hours work and at the end of the shift. As workplace locations are open to the elements we also collected ambient temperature readings to further assess their impact as a contributory factor to performance impairment. We identified that in cold temperatures < 0oC individuals have slower reaction times, as temperature increases reaction time decreases. Furthermore we identified that reaction time varies across the daily monitoring periods; generally following the slow-fast-slow relationship as found in our preliminary study. We also noted large variations in reaction time for individuals working in close proximity to mechanised plant, suggesting that the extra vigilance needed to complete tasks in operational areas that are shared with mechanised plant can negatively impact the speed at which an individual can react. The use of automated activity collection facilitated by the use of activity monitors has also provided an insight into the physiological demands of typical roles performed within the harvesting operations. We also suggest captured data from activity trackers can be used in conjunction with current risk assessment processes to add an extra dimension to the risk assessment process, facilitating a more informed view of individual role demands