Current developments in greenhouse robotics and challenges for the future

Horticultural production faces a growing demand for mechanization, automation and robotics as labour costs are increasing and the availability of skilled labour is decreasing. The modern consumer demands guaranteed and constant quality. Moreover, there are intensified hygiene, food safety and traceability demands. Automated production and quality assessment systems can contribute to fulfil these demands. Many highly automated systems are already applied in commercial high-tech greenhouses. This includes logistics and autonomous transport of plants and harvested product from the greenhouse to the sorting and packing facilities, spraying robots, machine vision based sorting systems for pot-plants and cut-flowers and robotic cutting, planting and grafting machines. Current research focus on automated crop scouting (e.g., insect and disease detection), phenotyping (e.g., monitoring and predicting fruit setting) and postharvest quality assessment. Many research activities are currently carried out on robotic systems that are capable to autonomously harvest crops and to conduct other repetitive and labour intensive operations. In most of these applications the success rate and cycle time are currently still too low to allow implementation in commercial practice. Human-robot collaboration and “human in the loop” applications are important stepping-stones towards full automation. These applications must also include the key issues of user safety and user acceptance. From the plant side, breeding crops with novel phenotypes and plant architectures, such as fruits which are easy to see and reach by robots, will simplify and accelerate the application of robotics in the horticultural field. Challenges involve the complex integration of multiple disciplines and technologies (e.g., navigation, safe operation, grasping, manipulation, perception, learning and adaptation). Current developments are supported by the worldwide rapid improvements in computing and sensor hardware, software and artificial intelligence