Regarding to real-time or online technologies in recent years, new technologies has been introduced into practical farming especially in the field of nitrogen application. These technologies are based on sensors mainly detecting the canopy reflectance. In the field of plant protection, although few sensor-based real-time technologies in weed control and growth regulator application are marked available, solutions for fungicide application are mostly missing currently. Amongst others at the Leibniz Institute for Agricultural Engineering (ATB) sensors for real-time plant protection are developed. The CROP-Meter was a first mechanical sensor for precise fungicide application in cereals that was available at the market. The sensor signal was correlated with plant surface and biomass which serve as parameters to vary the application amount by letting the concentration of the spray liquid constant. In the sensor-controlled treatment of long term field trials no yield reduction and no higher occurrence of plant diseases have been found in comparison to a common uniform treatment.

The operation of sensors, which work contactless, is more easily compared to sensors as the CROP-Meter which is in contact with the crop while spraying. A joint research project financed by the German Federal Office for Agriculture and Food (support code: 2814704511) was started in fall 2012 to develop real-time application technologies which uses non-contact sensors for precise fungicide spraying in winter wheat.

The joint research project consists of three subprojects:

The decision support system proPlant expert.classic or the internet-version proPlant expert.com (proPlant Co.) resp. delivers the basic principles of the Precision Farming Module “Fungicide” which will run on the onboard terminal in the tractor`s cabin to control the spraying process. The module will include 8 important leaf and ear diseases of winter wheat. After the suggestion of appropriate fungicides and their dosages the module defines the local target application amount while spraying by using the local sensor value as input parameter.

The ultrasonic sensor (Agri Con Co.) sends out short ultrasonic pulses. They are reflected by the leaf layers of the crop as well as the soil. The time-of-flight of the different echoes are measured. A “ultrasonic value” is calculated which correlates with crop height and biomass. The aim of this subproject is to find a calibration for different growth stages and varieties of winter wheat. One sensor will be attached to each section of the spray boom. Each section will be controlled separately.

The multispectral camera-sensor (ATB) takes red and infrared images simultaneously. These two images are used to calculate a grey scale image of the NDVI vegetation index. For separating green plant tissue from background a threshold is used to get a binary image. The percentage of pixels representing “green” gives the coverage of the green crop and represents the sensor value. This is also the advantage of camera-sensors. They separate crop and soil compared to optical sensors based on spectrometric measurements which are a mixture of crop and soil. Based on field trials in winter wheat the correlation between the sensor signal (coverage level) and plant surface (Leaf Area Index, LAI) as well as biomass is analyzed. In this subproject a variable rate spraying controlled by the camera sensor will be realized.

The joint research project is mainly divided in three main parts:

First results of the project especially from the first year field experiments 2013 are presented.