1.1 Sensor-based detection of mycotoxine producing fungi in wheat crops
Products made from cereal grain can contain toxins originating from fungal ear diseases. In wheat, the most occurring toxic ear pathogen is head blight caused by Fusarium spp.. Mycotoxins are dangerous in human and animal nutrition. An automatic detection in the field by machine vision could help to obtain information about the severity of the disease infection very rapidly. Before the harvested product goes into storage or is processed further, monitoring results could be used to decide whether the grain is to be used for human or animal nutrition or should even be put aside. Digital image analysis would offer a non-destructive, non-contact and fast tool for recognizing diseased ears. Digital cameras as on-line sensors can be carried on tractors and may be operated at common speeds of agricultural machines. In the project hyperspectral, thermography, chlorophyll fluorescence and colour images are to be tested on artificial infected wheat plants in the laboratory but manly under outdoor conditions in farmer’s wheat fields. A high wheel tool carrier will be equipped with a mechanical system on which different cameras are fixed. A simple and straightforward algorithm is to be developed to fuse the spectral information to detect diseased ears. Research work is focussed on the following two main problems:

• Evaluation of the detection reliability
• Determination of the specific date of the epidemic of the disease, when symptoms are detectable (time series).

1.2 Indicators and sensor technology for the identification of mycotoxine producing fungi in the processing of grain 
Module “Spectroscopic methods”
Laser-induced fluorescence (LIF), diffuse reflection (DR) spectroscopy as well as ion mobility (IM)spectrometry will be tested and applied for the detection of mycotoxin contaminations on grain, e.g., in storage facilities, during loading, or before processing.
In order to evaluate the capabilities of each method basic parameters are determined first (for model systems and in a subsequent step for real samples). Topics of the project are:

• Spectral range for DR spectroscopy (UV – NIR range) and characterization of suitable absorption and reflection bands for the recognition of mycotoxin contaminats on grain.

• Excitation and emission settings in LIF for the detection of mycotoxin contamination, either of the mycotoxins or fungi, respectively. The application of time-resolved detection schemes is tested in order to minimize the possible influence of signals originating from the complex background matrix.

• Characterization of potential application of IM spectroscopy for the detection of mycotoxin contamination in dust/air of grain storage and processing facilities.

• Possibilities of locally resolved detection schemes (e.g., for storage facilities) are evaluated (e.g., in combination with fibre optics).

• Determination of the limit of detection (LOD) for DR spectroscopy, LIF, and IM spectrometry.

Selectivity, sensitivity, and cross sensitivity are characterized for the different spectroscopic methods. For data evaluation advanced algorithms are developed. The ultimate goal is to built a mobile spectrometer system for the detection of mycotoxin contamination that can be used in-situ directly at the storage and processing facilities.

1.3  Development of a microwave-based humidity sensor for process control in grain dryers
Objective of the project is to develop an in-line microwave grain moisture sensor and to technically realize a model-based control concept for grain dryers. It is aimed to avoid the formation of mycotoxins as well as thermal damages and sales losses due to over-drying. In the phase of sensor development laboratory measurements at different types of corn are necessary. An existing mathematical model for the heat and mass transfer will be extended to continuous grain dryers. For this purpose, an approach for the solids transport must be derived and introduced accompanied by residence time measurements. To validate the model drying experiments at the ATB dryer test station have to be carried out. The microwave sensor and the control system will be proved both in pilot and industrial scale.

1.4 Technology assessment for sensor application in the production chain of grain
Chain-comprehensive technology assessment for estimating the socio-economic and ecologic efficiency is task within a superordinated subproject. In terms of technological impact assessment it aims to measure, analyze and evaluate the effects of sensor application at all critical points of the cereal grain chain. These effects may comprise e.g. modification of production processes, improved product quality and safety, reduced environmental impacts, enhanced rentability, and ameliorated working conditions. Both producers as well as users of the sensor integrated systems will benefit from the project outcome. Results will provide planning criteria for the implementation of sensor techniques into production chains and for the necessary adjustment of processes according to the sensor-induced changed conditions. Moreover, it will grant knowledge on the economic and ecological effects of the new technologies.

2.1  Assessment and evaluation of risks of quality loss in perishable horticultural produce
The project aims at development of a method to determine the quality and to evaluate the spoilage risk of horticultural products for use close to the production chain. In cooperative work together with a SME, a spectral sensing device for produce quality will successively be developed and tested. The evaluation is based on determination of quality at the beginning of the production chain as well as on representative data of mechanical and climatic stress during run of produce through the chain, acquired before by means of temperature data logger and a miniaturized sensor for mechanical impacts developed in frame of previous research project. In this way, the decision for optimum utilization of produce charge will be possible at an early stage. Development and tests of spectral sensing device will be carried out under laboratory and practical conditions. Several years tests and optimizing measures are planned, in order to meet the demands on sensing device and on evaluation algorithms for characterization of produce quality and its change due to influencing factors in the production chain. Results should be useful to improve the quality management. They are basically for further development of guidelines for certification of production technique.

2.2  Development of biosensors for the detection of human- or phytopathogenic microorganisms in the post harvest chain of perishable products: I. Cytometry-based analysis, II. Molecular characterization
After harvest vegetables and certain fruits are contaminated with basting substances like soil particles and organic residues. Depending on the type and the degree of contamination, varying concentrations of both human- and phytopathogenic microorganisms can emerge. Thus, for processing fresh produce, adequate washing steps are indispensable. The consumers demand for food safety and product traceability increases rapidly. Therefore, one major task of the research project is the estimation of microbial load during food processing and the consequent monitoring of the hygiene status. However, conventional microbial investigations are often time-, material- and work-consuming. Additionally, in most cases microbiological methods are insufficient to differ closely related bacteria species. Thus, classical microbiological techniques are inadequate for required real time monitoring steps in food processing procedures. Within the context of this project new tools for specific and quantitative detection of spoilage microorganisms shall be developed, which will allow a continuous monitoring of microbial contamination in washing steps during industrial processing of vegetables. The project focuses on the application of modern molecular approaches like the flow cytometric quantification and differentiation of bacterial cells supported by PCR- and realtime-PCR-based genotyping. The objective is the estimation of the basics for simple automated system (“biosensor”) for the specific quantification of spoilage organisms. A utility model shall be constructed and tested for optimization of washing procedures in post-harvest processes.

2.3  Modular system for quality monitoring in the logistic chain of perishable horticultural produce
Intelligent systems are needed to meet the progressive requirements on quality, safety and traceability of fresh horticultural produce without great additional personnel effort. Developing an intelligent system which is able to completely supervise and to evaluate meas-urements of quality relevant parameters from the producer via transportation and short-term storage up to retail is the main objective of the project. Modular datalogging units for the measuring of air temperatures and air humidities (and for the assignment of text information) which can be used both self-sufficiently at the lower level and in a system based on internet technologies, will be developed in a cooperation between the Leibniz Institute for Agricultural Engineering (ATB), the Federal Institute for Materials Research and Testing (BAM) and ESYS GmbH, Berlin. Taking the fruit and vegetable production as an example an essential contribution shall be made to improve the sustainability in the supply chain by the target-oriented application of the system. An important part of the project to be developed is the modelling of the postharvest behaviour of harvested fruit and vegetables in different packaging/transport units under changing environmental conditions.