The grow-out house service robot (GOHbot) is designed to perform maintenance within a chicken house in a more sanitary way than that of human maintenance. The purpose of GOHbot’s implementation is to improve flock quality by reducing bacterial contamination without compromising meat yield. The literature discusses disease prevalence in chicken grow-out houses, especially that of Campylobacteriosis, which requires a solution like GOHbot. Multiple sources discuss the unique chicken grow-out house environment and the challenges it presents for the maneuverability and perceptive capabilities of agricultural vehicles like GOHbot. Several sources outline the technologies that govern LiDAR scanners and Kinect 3D cameras, both …show more content…
In addition to corroborating that the coarse terrain on which GOHbot operates as a challenge to obstacle detection and therefore maneuverability, Sherman (2013) details the navigational obstacles that are prevalent in grow-out houses, citing that a broiler flock contains roughly 20,000 birds in a grow-out house about 400 feet long and 40 feet wide. This housing provides roughly 0.8 square feet per broiler once they eventually grow into the space (Nedrick, 2012).
In contrast to the previous arguments, Sherman (2013) points out constant variables within grow-out houses, as temperature and humidity are managed by changing ventilation either naturally by adjusting curtains along the house’s sides or through “tunnel ventilation” using vents and fans. Since no literature seems to regard temperature and humidity as possible confounding variables for obstacle detection, it is fair to assume that GOHbot’s obstacle detection should remain unhindered by variations in atmospheric conditions.
Obstacle Perception in …show more content…
According to Fankhauser et al. (2015), there is a general consensus that the first version of Microsoft Kinect is a depth sensor that quickly returns high-resolution depth images for a low cost. The original version of the Kinect 3D camera uses a real-time algorithm to calculate distance by triangulation using randomly-projected infrared light beams (Chen, Yue, Wu, & Wang, 2014), and presents point-like depth readings in the form of point clouds (Stefanczyk et al., 2013). According to Fankhauser et al. (2015), the second version of Kinect is a time-of-flight depth sensor, meaning that it implements a detection process in which obstacles reflect a strobing infrared light and the infrared camera registers each pixel's time of flight to the obstacle and