Smart Agriculture IoT Living Lab

This use case is expected to demonstrate significant benefits arising from IoT exploitation in optimizing various aspects of smart agriculture. The involved IoT devices include multiple types of ground micro-climate, soil and leaf information stations, drones, mobile robots, as well as wearable devices. In this use case, IoT-NGIN will be piloted against several agricultural goals:

  • to enable the assertion of the evolution of the crop
  • to detect diseases
  • to optimize irrigation and fertilization
  • to reduce spraying and support manual fruit harvesting

Moreover, data from the micro-climate may be used for other applications, e.g. identifying the driving conditions in nearby roads and highways.
The use case will be hosted at a commercial orchard in the region of Peloponnese, Greece. The pilot site has a size of 300 ha and belongs to Cooperative Winery of Nemea, a pioneering company in the field of winery production.

Technologies already in place include an anti-frost system, a hail protection system, automated irrigation and a fertilisation system. Monitoring the status of the growing crops is a very important factor for getting decisions on irrigation, fertilization and harvesting, which in many cases, significantly influence the quality and the quantity of the crop. As such, the SynField system by Synelixis (IoT nodes and cloud platform) has already been installed in the Nemea orchard.

Use case Applications

  • Crop diseases prediction. Smart irrigation and precision aerial spraying.Crop diseases can be in many cases predicted or early detected using micro-climate measurements (mainly temperature and humidity at the air, the leaves and the soil), crop image processing and visual analytics. Within IoT-NGIN, beyond micro-clima measurements utilising the Synelixis SynField precision agriculture IoT nodes, we will also experiment on crop diseases prediction with images and real-time video analysis of the crop and the leaves captured from visual and multi-spectral cameras located on semi-autonomous drones flying over the orchard. By performing video analysis in real-time either locally (on the drone), based on already trained ML models, or remotely (at the edge) based on federated ML, the drones will be able to dynamically modify their trajectory to introduce optimal, precision aerial spraying only in areas of interest. Moreover, utilising the IoT-NGIN technologies, orchard micro-clima data will become available to third parties.
  • Sensor aided crop harvesting.Open-air horticulture is a labour-intensive operation that accounts, in many crops, for about 50% of the total production costs. Though experiments with apple harvesting robots have been introduced, fully automated machine-based harvesting systems in vineyards have not yet been effectively applied. Hence, knowledge-based synergetic mechanization is proposed as an intermediate option: Automated Guided Land Vehicles (AGLV) that will support human workers by autonomously carrying the crates to the loading point. BoniRob, Bosch Agricultural Robot, is a good candidate for AGLV for this use case. We will experiment with agriculture AGLV serving as carrier machines, by enabling them to locate and avoid workers (for safety reasons) and trees (for operating reasons). The correlation between crates and trees can be automated using RFID tags on crates, while long-range RFID readers may be mounted at the loading points. Wearable devices can collect location data and workers’ activity data.

IoT-NGIN as Precision Agriculture IoT Differentiator

  • AGLV-humans collision prediction, detection and avoidance through real-time positioning and federated ML functions hosted locally on the IoT nodes and in edge computing
  • Combine edge (remote) and semi-autonomous (local) federated ML decision on constrained resources, enabling dynamic drone’s trajectory modification for optimal spraying and safe operation of AGLVs in the field even in case of network error, loss of connectivity or weak coverage
  • Contextual-IoT to inform the AGLVs on the location of crates and loading points