Research Projects

Smart agriculture relies on collecting and analyzing farm data to enhance crop yields, optimize resource management, and improve farmers' quality of life. In India, the adoption of IoT-based devices is increasing, leading to the need for cost and energy-efficient solutions. Duty-cycling is commonly used to reduce power consumption, but it often leaves the IoT node off most of the time, hindering the capture of frequent or irregular events. To address this issue, always-on event detection should be used close to the sensor, allowing the rest of the system to be kept off, saving power. The proposed solution involves developing an ultra-low power sensor interface called an analog front-end (AFE) with event detection capability. The AFE acquires analog output signals from sensors, then digitally processes them (DSP) and wirelessly transmits them (TX). Circuit techniques to operate MOS transistors near cutoff regions can reduce power consumption to the picowatt level, allowing always-on operation of the AFE. The project also generates wake-up triggers for battery-powered DSP and TX, waking subsystems only when monitored analog signals exceed certain threshold limits. This minimizes IoT power consumption and prolongs battery life. Once activated, the DSP performs data analysis using low-complexity machine learning and transmits the decision. In summary, the project scope includes identifying key soil condition parameters, developing an interfacing analog/mixed-signal integrated circuit for ultra-low power signal acquisition and event detection, and designing cost and energy-efficient microcontroller-based back-end processing and data transmission.