Internet of Things cultivation boosts greenhouse horticulture
First Consulting has collaborated with Hortilux Schréder, a supplier of professional grow light systems for greenhouse horticulture, to set up a model to predict the expected fresh weight production of greenhouse crops. With the help of practical knowledge of expert gardeners, variables have been selected that influence the fresh weight production (kilograms of tomatoes or roses). Hortisense applies machine learning to create a forecasting model.
This scientific growth model calculates how much sugar (assimilate) produces a crop, based on input variables such as light, temperature and CO2. By combining sensor data from the growth light systems supplied by Hortilux with this physiological knowledge, useful input variables are created. Subsequently, machine learning algorithms can be applied that summarise crucial relationships between variables in a model. In addition to a prediction, the model also provides insight into the impact of these input variables on fresh weight production.
Commenting on the future of the project, First Consulting agile development practice lead, Andre van Teeseling, said, “Three years ago, this was just an idea. Hortilux knew what they wanted to achieve initially and shared this with us, but we recognized from the start this was the beginning of a journey. And we were right, because we are still learning and adapting Hortisense as we go along.”
Read more at Consultancy.uk