he domain emphasises on sustainable data collection with affordable sensing infrastructure, data analysis with state of the art AI/ML methods, and distribution of the information with an adaptive human-computer interface for decision support and task automation.
Receiving information is the key to a cyber-physical system, and particularly important for crop surveillance and monitoring systems. The sensors capture environmental, crop, and activities taking place on the farm. The goal of this project is to decouple the sensing infrastructure from farms and provide surveillance as a service using modern sensors such as UAV, Satellite, Robots, etc.
Due to the enormous spatial size of the farmland and regular interval of data-sensing makes the big data analytics challenging and costly. The overall cost includes the cost of storage, transmission, and computing resources. The goal of this project is to build affordable AI/ML solutions to help farmers in agriculture automation, decision support, precision agriculture, and market analysis.
Farmers are not technology savvy. Most of the farmers find it difficult to use technological devices. In developing countries, farmers may even not be familiar with the English language. The aim of this project it to build human-computer-interface (HCI) that are farmer-friendly. Particularly we intend to consider the local culture in choosing the interface modality and structure.
Due to the vast fields, replacing batteries in sensors, wireless devices, and micro-controllers is a challenging task. Therefore, it is beneficial that battery requirements are minimized in computational and sensor devices deployed on the field. However, the state of the art of AI/ML algorithms have a huge computational requirement. Therefore, we aim to build custom electronic hardware solutions that are cheap efficient, specifically for precision agriculture and monitoring tasks.
iHub – AWaDH is established by the Department of Science & Technology, Government of India, at the IIT-R in the framework of NM – ICPS with time-bound predefined deliverables.