Development of a low-cost traffic counter for assessing likelihood of impact for tree risk assessment

Abstract

ABSTRACT Most tree risk assessment methods task the assessor with the evaluation of a tree’s likelihood of impact (i.e. likelihood a target is present during failure) when determining the probability that an adverse event will occur. While this is generally accomplished qualitatively based on visual cues; during the 15 to 20 minutes an assessor is at a location, site occupancy, a key component of likelihood of impact, is most accurately measured using traffic counters. The use of traffic counters reduces risk assessment bias and increases reproducibility, though commercially-available devices may be cost-prohibitive. This work addresses one potential barrier to traffic counter adoption through the creation of a low-cost vehicle and pedestrian counting system. The study compares the accuracy and consistency of our homemade traffic counting system to estimates derived from a commercially available system, as well as vehicle and pedestrian counts tallied with a hand clicker. Vehicle and pedestrian counts from the commercially available systems (p = 0.004 and p < 0.001, respectively) correlated well (r = 0.846 and r = 0.896, respectively) with the hand-tallied counts, but the homemade counters were not correlated with either. However, pedestrian counts from the homemade counter became inaccurate as afternoon temperatures went above 31.7°C.