Academic Innovations

A digital twin for smart agriculture system, which can emulate various network attacks, progression, and impact of the attacks. This will aid in designing prevention systems, such as optimal placement of the attack detectors, and design of the IoT systems with sensory hub, and actuator hub sub-systems. The actuator sub-systems control the release of water and water-soluble nutrients. Additionally stationary wide-angle imaging system will be deployed to periodically update the health of the IoT subsystem, growth of the crops around its view, and detection of any pest-attacks using the imaging system. Demonstrating a stationary 2-DOF imaging system to periodically update the health status of the IoT hub and simultaneously providing feedback on overall growth of the open farmland and controlled (poly) house where it is deployed. Detection of bio-attacks using the imaging system will be established for the crops near its vicinity. Veracity of the results will be validated using low-cost mobile robots (or drones). This includes various pest attacks during the growth which are categorized for focused manual treatment post detection. Demonstrating usage of low-cost robots (or drones) for the overall surveillance of the IoT subsystem and imaging system. Further a possibility of collecting ground truth collection of sensor data parameters. Parameters such as temperature, humidity, soil-moisture, and NPK can be measured and used to verify the accuracy of credibility of collected data from the IoT subsystem. This also helps in detecting security breach along the IoT subsystem. A policy framework recommending actions for different attacks will be developed and the same will be followed in the deployed system. Any kind of an attack will be classified to either global or local effect based on its connectivity and impact. In the category of local upset the node will be switched off to disable any data collection from that subsystem. In the global category, the system will be suspended till a complete analysis is performed and further switching on the system on priority basis is expected. The policy draft will be submitted based on our experiments and analysis. A policy framework recommending actions for different attacks will be developed and the same will be followed in the deployed system. Any kind of an attack will be classified to either global or local effect based on its connectivity and impact. In the category of local upset the node will be switched off to disable any data collection from that subsystem. In the global category, the system will be suspended till a complete analysis is performed and further switching on the system on priority basis is expected. The policy draft will be submitted based on our experiments and analysis.