A preliminary study of remote control skidder operation
Type of DegreeMaster's Thesis
MetadataShow full item record
In timber harvesting, skidders are widely used by operators who are in charge of pulling the woods. However, operators have to visually monitor multifarious objects such as loggers, machinery, trail, terrain and trees are being cut for the sake of work plan and personal safety. Since distraction may lead to poor efficiency, injuries or even death, it is indispensable to determine not only how the operators handle the skidder but also what they look at along with the distribution of their attention and why they concentrate on specific area under relevant task. Three GoPro cameras were responsible for the 3D orientation model of the gaze. The first GoPro camera was attached to bracket in front of the hard hat in order to get the frame of the gaze. The second GoPro camera was mounted on the back of the cab so as to capture the head motion during forward looking. The third GoPro camera was then installed on the left side of cab so that the head action during backward looking can be tracked. A computer was then used to process the image, restore the view of the operators, calculate the density of the gaze behaviors and generate a heat map of the gaze overlaid by a virtual perspective. In order to reduce the accident rate and fatigue caused by the long time in-cab timber harvesting operation, as well as improve the efficiency such as multiple tasks execution, the video-based driving system built for testing the feasibility of remote driving was set up in the cab. A fisheye camera was installed on the head of the skidder, a monitor was mounted on the region over the steering wheel, and a laptop which was in charge of process the video was connected with the cameras and the monitor. Also, the GPS information like location and speed was recorded. The results indicated that the remote control based on the live video was feasible.