A robot that can help forecast tomato crop yields by autonomously measuring the size, quantity and ripeness of production in large-scale greenhouses could be ready for commercial launch next year, Fruitnet has learned.
The fully automated Plantalyzer, which is currently being trialled in the Netherlands by the Greenhouse Horticulture Business Unit at Wageningen University & Research (WUR) – with the backing of HortiKey, part of Berg Hortimotive – could eventually enable growers to create yield maps of their entire greenhouse.
By employing the data generated by daily scanning, those growers would in theory be able to compile detailed forecasts predicting volume as well as quality.
In the future, the machine may also be used to identify signs of disease by picking out spots on the plants' leaves or fruit.
Plus, once it has been perfected for tomatoes, the plan is to develop a version of the system for cucumbers.
'We are still at the beginning but we are trying to make a good forecasting solution that can be brought to market next year,' comments Erik Pekkeriet, senior business development manager for agro food robotics at WUR.
'The goal is that by 2020 we will be able to use the system on a full greenhouse, monitoring yield, size and ripeness to get a good idea of the count and the size of tomato crops.'
Initial tests on a self-propelled Plantalyzer prototype began in 2018. Since then, the number of cameras used on the machine has been increased from two to six.
Arranged in three pairs, those cameras can build a detailed picture up to a range of 1.5 metres from the plants, broadening the robot's field of vision and therefore making it potentially more suited to being used in large-scale industrial greenhouses.
Software developed by WUR then analyses the images and data sent by the robot via wifi to a central computer, where each individual fruit is compared to a so-called harvest card of variety-specific reference colours for each stage of production.
