Development of a novel coconut-tree-climbing machine for harvesting

Abstract

Presently, there is a dearth of professional climbers to carry out coconut harvesting because it involves considerable risk posed by the tree height and uneven trunk surface. Numerous coconut-tree-climbing devices have been developed to overcome this issue, but human effort is still required. Hence, it is imperative to develop a wheel-driven mechanism in the structure with a continuous motion to harvest coconuts. By featuring an anti-falling capability, the climber can be stable in static and dynamic conditions. This feature can be achieved by attaching springs in the supporting frame during vertical climbing on the trunk surface. This paper presents a 3D modeling and the static and dynamic analysis of a climbing machine. The climbing field trials of the climbing model were carried out on a coconut tree trunk, and the best climbing rate that was achieved was 0.3 m in 1.6 seconds. Control of the climber maneuverability was tested with an embedded processor board. The maturity identification of coconuts in a complex background was successfully achieved with a detection score of 99% on real-time images by using deep learning techniques. This study can be extended to identify the cutting point of the coconut bunches for harvesting.