Agriculture has undergone a transformation because of the combination of the ‘ML algorithms’ and the ‘IoT’ technologies, especially in the sector of ‘disease diagnosis’ in crops like maize. In this paper, it discusses about recurring problem in agricultural production by presenting a preferable technique that takes use to reliably detect diseases in maize leaves. IoT sensors capture data in real time, which is then analyzed through ‘machine learning algorithms’. This paper improves the accuracy of illness diagnosis with the ‘machine learning techniques’ and a comparative analysis of detection accuracy at different levels. The average accuracy of batch size 32 at 60 epoch level is resulted as 93.80%. Similarly at 30 epochs resulted as 92.19% and at 10 epochs it was 90.18%. In addition to being technically novel, our concept has significant ramifications for food security, cost-effectiveness, and agricultural sustainability. Farmers may lessen crop losses, lessen their need on chemical treatments, and eventually promote more resilient agricultural methods by detecting problems early and treating them specifically and protect the world’s food supply.