Collaborative smart-robot for yield mapping and harvesting assistance

Abstract

Since the quality and demands of agricultural products increase, there is a growing need for human labour while increasing yields and reducing costs. For example, one specific case where the harvesting tasks remain mandatory manual is the vineyards with table grapes or grapes destined for high-quality wine. Therefore, this paper presents an operator tracking strategy for a mobile, collaborative robotic platform. The system applies to manual fruit picking. It addresses human-robot collaboration while harvesting. The aim was to develop a harvest assist robot to carry the grape box and would also be able to generate an accurate yield map, based on information provided by a GNSS-RTK receiver and a scale. Both devices were on board the robot. The results showed that the robot successfully performed the operator tracking process, reducing the physical effort required in harvesting and maintaining product quality, while mapping yield with an approximate error between 0.25 and 0.3 kg.