Automation in Farming with IOT

Internet of things (IoT) is an emerging technology that shows the fu- ture of computing and networking. Agricultural monitoring from a remote location is one of the essential applications of IoT based wire- less sensor networks. The IoT based wireless sensor network faces problems due to the dynamic changes in the environment. The num- ber of required sensor nodes increases for monitoring of the vast area. By introducing mobility of all nodes in the IoT based wireless sensor network, we can decrease the number of nodes and thus re- ducing the cost of the overall system. In this research project, an IoT based mobile robotics network is proposed for farming applications. Master and slave robots incorporate the wireless sensor network and are connected via the NRF protocol for reliable sharing of sensor data. The master robot also transmits this data to the IoT server. The highlighting features of this research include weeds detection through image processing and sensors for gathering light, moisture, humidity, temperature parameters. Robots are also equipped with an ultrasonic sensor for avoiding obstacles during navigation. The proposed computer vision algorithm for weed detection is based on texture feature analysis and artificial neural networks. The algo- rithm is implemented on Raspberry Pi 3 based single board com- puter for classification of weed and non-weed images.