Energy consumption in greenhouses

The Challenge:

One of our customers in the Agritech sector builds and operates Greenhouses for the food industry in harsh environments in the world. With roughly 500,000 m2 of Greenhouse space spread over 3 continents the company guarantees a minimum yield of tomatoes, bell pepper, cucumber and alike.

Temperature, Carbon Dioxide and relative humidity control in these extreme world regions is critical; a 5% saving on energy consumption can easily translate into 2% improvement of bottom line result. The greenhouses are controlled by a semi-automated Operations Control unit for the local operator in Kazakhstan or China to monitor the temperature and humidity input through the Air handling units. The company itself, located in Europe, needs more detailed output-related measurement to see the actual impact of Cooling or Heating on specific zones in the Greenhouse where the crop grows.


The Solution:

FS2D offered the company a COOLBOX with a number of targeted sensors to measure water temperature and motor current of the Air handling unit, and sensors to measure air temperature and humidity for several zones in the Greenhouse. Hooked up to the INSIGHT Cloud platform through the Router with 4G and secure VPN-connection, the Company can monitor precisely the impact of the Operations Control unit on crop maturity. They found that the unit turned up the heat in the greenhouse either too high or kept it too low, thereby consuming unnecessary energy. By adjusting the parameter settings in the Air handling unit and provide output feedback to the unit, energy consumption reduced with 7% and temperature readings were no longer off the charts.

Next step is to link the INSIGHT platform to Meta data like weather forecast and Airpressure predictions to anticipate changes in outside temperature and solar radiation to reduce energy consumption even further. Further out, the currently captured data will be extended with vision technology. The collected, benchmarked and interpreted data is input to build a machine learning model with a constant feedback loop of crop growth, sensor readings and parameter adjustments in the Operations Control units in all of the company’s Greenhouses across the Globe.


The Reward:

The Greenhouse Operator has been able to reduce energy consumption and improve crop-yield, resulting in improved revenue and increase in bottom line results. They will further improve yield across their install base with the Machine Learning model currently under development.