ANTARES: advanced technologies for sustainable agriculture and food security

Photo: Zoran Cvetković. File: Benediktanski manastri Arača, Novi Bečej 13.JPG. This file is licensed under the Creative Commons Attribution-Share Alike 3.0 Unported license.

The overall aim of ANTARES is to transform the BioSense Institute of Novi Sad (Serbia) into a European Centre of Excellence for Advanced Technologies in Sustainable Agriculture and Food Security. BioSense is a leading research institute in ICT which wants to apply its expertise to agriculture and food in collaboration with Wageningen University & Research (WUR). WUR is a global leader in agricultural and food-related research with a strong presence in EU-funded research and strong relations with private industry.

Synergy by combining expertises

The synergy realized by combining the specific expertise of BioSense and WUR can be used to solve problems that have hitherto been unsolvable. Through ANTARES, BioSense will benefit from the fact that WUR is a leader in agrifood research and well-placed to articulate the need for ICT in agriculture and food. WUR will benefit from the highly complementary expertise of BioSense in the field of ICT, sensor development, and data mining.

In the framework of ANTARES, exchanges will take place of researchers, technicians, and supporting staff, from both institutes.

Examples of synergies that can realized by combining ICT and agricultural science:

Phenotyping

WUR breeders, geneticists and statisticians work on plant phenotyping, which aims to understand how a plant’s genotype leads to its phenotype. This approach is leading to new varieties with high and stable yields and with a small environmental foot print. BioSense scientists will strengthen ongoing work by creating nano/microelectronic sensors for measuring plant phenotypes and by developing big data analytics to make sense of the terabytes of information generated by today’s high-throughput phenotyping facilities.

Yield prediction

A farmer’s decision on how much fertilizer or pesticide to use, is typically predicated on the yield he or she expects. The WUR approach to yield prediction, which is based on understanding and modelling the processes leading to yield formation, is not necessarily well-placed to make efficient use of the large amounts of data that are gathered today by automated crop and soil monitoring. BioSense has already successfully predicted yields using data mining techniques, without, however, taking into account the all-important influence of weather. By combining insights from crop physiology on the part of WUR scientists with a deep understanding of artificial intelligence on the part of BioSense scientists, we will be able to make highly accurate yield predictions and precise fertilizer and pesticide recommendations.

Decision-support

Precision Agriculture is a cyclic system where information about crops, soils, weather and management is recorded continuously (daily, monthly, yearly) and used for subsequent decision-making. WUR scientists know which information is worth recording and how to make use of it in simple decisions. BioSense scientists know how to build and operate the necessary sensors and how to extract previously unknown patterns from the vast amounts of data collected in this way. By working together, WUR and BioSense can achieve increases in resource use efficiency that neither institute would be able to achieve by itself.

Big data modelling and knowledge modelling for smart food chains

Using advanced robots and sensor systems, agro-food processes can be monitored throughout the chain. From remote sensing of the crop with drones, to autonomous harvesting and automated post-harvest quality assessment of the produce. Relevant (quality) information will be extracted from the sensor data using deep-learning methods. The information gathered in the chain will be coupled, allowing the discovery of patterns and future predictions based on big-data analysis and allowing personalized communication to consumers.

In the “E-Pieper” public-private partnership, WUR and BioSense will develop a multi sensor system that gives field and site specific information on conditions in the soil where potatoes are growing, plus yield information of stored potatoes. This could be a sensor package that looks like a potato, is planted with the potatoes, and is “harvested” when the potatoes are harvested. Climate sensors for storage and micro-climate in field are already available, but we miss sensors for soil properties and sensors for yield and quality of harvested product. The sensor data will be combined with crop growth and management models, in order to deliver strategic and operational advices.

A result of the WUR- BioSense collaboration, in the near future, could be drones autonomously monitoring an apple orchard. They assess information about the growth and health status of individual plants and the quality of soil over time. When the apples are predicted to be ripe, (semi-) autonomous robots harvest the fruit allowing tracking and tracing of in the downward chain. An automated quality assessment system then accurately checks the quality of every individual product for optimal sorting, logistics and storage. By coupling the data throughout the chain. Patterns can be observed linking the drone data for the quality data of individual fruits. These patterns can help to predict yield in the future and to optimize farm management. The data can furthermore be presented to the consumer, allowing applications for personalized nutrition.

Springboard for Dutch companies to Balkan

ANTARES will be of interest to Dutch companies active in Serbia, Croatia, Bosnia-Herzegovina, Montenegro and Macedonia. These countries are of interest for Dutch companies as alternative production location; builders of greenhouses and producers of seed, seed potatoes and soft fruit are already active in the region.

ANTARES will create a demonstration farm in the vicinity of Novi Sad where companies can showcase their innovative agritech solutions. Farmers will be able to see solutions of interest to them implemented in real-world settings, while scientists from BioSense and other collaborating institutions will be able to run experiments and test prototypes in the operational environment. The demonstration farm will also be a place for regional stakeholders to meet, attend trainings and participate in co-development workshops.