Caracterisation des obstacles nuisant au travail des debroussailleurs et de leurs impacts sur la productivite par la modelisation et l'analyse du travail cognitif

Piece-rate payment systems (unit rate in ;Three steps were taken to reach that goal. The first step was to model productivity based on competing vegetation and obstacles abundance using linear regression analyses, and to explore the relationship between individual obstacle types and productivity through principal component analysis. Work studies were conducted in 0.12 ha treatment plots, with the participation of 91 subjects representing 22 silvicultural contractors. The best regression model (R 2=0.61, RMSE=4.3 hours per hectare, n=91) predicted effective time consumption (ETC) using three variables: 1) percent cover (%) of three brush species (raspberry, ferns, fireweed), 2) density of trees and shrubs higher than 1 m (number per hectare) and 3) density of obstacles hindering worker progression in the field (woody residue, stumps left after harvest, boulders and terrain roughness). Principal component analysis (PCA) was used to further revealed that two components representing woody residue and stumps bypassed by the worker were significant predictors of ETC (R2=0.14, RMSE=6.4 hours per hectare, n=91).;Measurements of hindering obstacle abundance have not been found to be easily reproducible, suggesting that some cognitive processes may be involved in coping with changing conditions. Forty-one participants were engaged in observations and interviews. Hierarchical task analysis (HTA) was performed to describe release work, complemented by an analysis according to the Skills-Rules-Knowledge (SRK) framework. Cognitive content seemed higher when workers were faced with obstacles. Half of the observed time was spent in obstacle-free areas while 30% was spent coping with slash and slash piles. Results should help develop a valid and practical method to account for obstacles in the field and may also help improve brushcutter training programs.;The final part of this dissertation aimed at testing the relationship between individual obstacle types and brushcutter productivity in the regeneration release treatment. Work studies were performed with the participation of 19 subjects employed by 11 different silvicultural contractors on as many sites spread across the province of Quebec. Brushcutter trajectories were tracked during the observation period using a Garmin GPS 60cx receiver. Exact start time of observed activities (brushcutting, filling up gas tank, filing the blade, resting) were tallied as well as obstacles that hindered brushcutter progression. A bullet camera was mounted on brushcutter safety hat to capture what was going on in front of the brushcutter. The video sequences were subjected to a time study using the Video Event Analysis software (Chappe Software) to measure the start time and duration of work around each type of obstacle as well as time in the absence of obstacle. The observation data and the video analyses data were synchronized with GPS trajectory data based on time (N=4 subjects). Calculated worker speed was contrasted between obstacle types using the SAS Mixed procedure. It was found that only slash and slash piles were significantly affecting walking speed by 41 meters/hour (p=0,0042) and thus lowered productivity. More than half of the observed work time was obstacle-free.