AgriSense: An AI and IoT-Based Smart Farming System for Crop Disease Detection, Multi-Farm Monitoring and Automated Irrigation

Abstract

Crop disease identification, field monitoring and irrigation control are available to farmers today only as separate tools: leaf-disease classifiers are not linked to field data, sensor-based irrigation controllers do not examine crop health, and farmer chatbots have no access to the farm. No single system combines these functions for several farms through one data path. This paper presents AgriSense, an integrated artificial intelligence and Internet of Things (IoT) system for smart farming. A MobileNetV2-based convolutional neural network classifies crop-leaf images submitted through a Flutter mobile application to a FastAPI backend and returns the crop, the detected condition and a confidence value, together with preventive measures, pesticide information and watering advice. An ESP32 controller reads soil-moisture, temperature, humidity and pH sensors, uploads the readings under a farm identifier and drives a relay-operated water pump with a hysteresis rule, with manual and timed operation as alternatives. A farm dashboard with scan history, alerts and a chatbot complete the application. The architecture, algorithms and prototype implementation are described, and a reproducible evaluation plan is given.

Introduction

The text presents AgriSense, an integrated AI and IoT-based smart farming system designed to improve agricultural productivity through automated crop disease detection, intelligent irrigation, and centralized farm management. Traditional farming practices depend heavily on manual crop inspection, disease identification, and irrigation control, which can lead to delayed disease diagnosis, water wastage, and difficulties in managing multiple farms.

Existing agricultural technologies generally provide separate solutions for disease detection, soil monitoring, irrigation automation, and farmer assistance. However, these systems often lack integration and do not share farm-related data. To overcome these limitations, AgriSense proposes a unified platform that connects crop health analysis, environmental monitoring, irrigation control, and agricultural recommendations through a common farm identifier.

The literature survey reviews 15 research studies covering IoT-based agriculture, CNN-based plant disease detection, agricultural chatbots, automated irrigation, MobileNet architectures, and edge computing. These studies demonstrate the effectiveness of individual technologies but reveal limitations such as single-crop applications, lack of multi-farm management, absence of integrated irrigation control, and limited connectivity between disease detection and real-time sensor data.

The research identifies four major gaps: lack of integration between agricultural technologies, restriction to single-crop or single-function applications, absence of centralized farm-wise records, and unreliable disease predictions without confidence verification.

To address these challenges, the proposed AgriSense system has six main objectives:

1. Develop an IoT-based field monitoring unit using ESP32 to measure soil moisture, temperature, humidity, and pH while controlling a water pump through a relay.

2. Develop an AI-based disease detection service using MobileNetV2 to identify crops and diseases from leaf images and provide prediction confidence.

3. Generate preventive measures, pesticide information, and watering recommendations based on detected crop conditions.

4. Implement automatic, manual, and timer-based irrigation using soil moisture measurements and hysteresis-based control.

5. Maintain centralized records of crop disease scans, sensor readings, and irrigation events for multiple farms using unique farm identifiers.

6. Develop a Flutter mobile application featuring a dashboard, historical records, alerts, farm monitoring, and an agricultural chatbot.

The methodology consists of two interconnected workflows. The first is the AI-based disease detection workflow, in which farmers capture or upload leaf images through the mobile application. The images are processed by a FastAPI backend and analysed using a MobileNetV2 deep learning model. The system identifies the crop disease, evaluates prediction confidence, generates relevant agricultural recommendations, and stores the results in the selected farm's history.

The second is the IoT-based monitoring and irrigation workflow. Sensors continuously measure environmental and soil conditions, while the ESP32 controller processes the readings and communicates them to the backend. Based on soil moisture levels, the system automatically activates or deactivates the relay-controlled water pump. Sensor readings and pump activities are also stored for future monitoring and analysis.

Conclusion

AgriSense unifies MobileNetV2-based crop disease detection, preventive and pesticide advice, ESP32-based sensing and relay irrigation, multi-farm management and a chatbot, functions that existing solutions provide separately, in one low-cost mobile platform. The farm identifier keeps images, readings and pump events organised for every farm, and the confidence check and the hysteresis control make the system cautious in both diagnosis and watering. Future work includes the quantitative evaluation of the classifier and of water saving described above, weather integration, additional soil and water-level sensors, on-device inference with TensorFlow Lite [14], [15], GPS-based farm maps, regional-language support, encrypted device communication and a solar-powered field unit.

References

[1] M. Ayaz, M. Ammad-Uddin, Z. Sharif, A. Mansour, and E.-H. M. Aggoune, “Internet-of-Things (IoT)-based smart agriculture: Toward making the fields talk,” IEEE Access, vol. 7, pp. 129551–129583, 2019. [2] M. Momaya, A. Khanna, J. Sadavarte, and M. Sankhe, “Krushi – The farmer chatbot,” in Proc. Int. Conf. Commun., Inf. Comput. Technol. (ICCICT), 2021, pp. 1–6. [3] B. Arora, D. S. Chaudhary, M. Satsangi, M. Yadav, L. Singh, and P. S. Sudhish, “Agribot: A natural language generative neural networks engine for agricultural applications,” in Proc. Int. Conf. Contemp. Comput. Appl. (IC3A), 2020, pp. 28–33. [4] J. E. Tamil Malar and M. Vaishnavi, “IoT based smart irrigation system using ESP32,” in Proc. 3rd Int. Conf. Electron. Sustain. Commun. Syst. (ICESC), 2022, pp. 1751–1755. [5] A. Salam, M. Naznine, N. Jahan, E. Nahid, M. Nahiduzzaman, and M. E. H. Chowdhury, “Mulberry leaf disease detection using CNN-based smart Android application,” IEEE Access, vol. 12, pp. 83575–83588, 2024. [6] R. Rashid, W. Aslam, R. Aziz, and G. Aldehim, “An early and smart detection of corn plant leaf diseases using IoT and deep learning multi-models,” IEEE Access, vol. 12, pp. 23149–23162, 2024. [7] J. Samuvel Joseph and G. Srinitya, “Deep learning based rice plant disease detection and classification system with automated treatment recommendation engine,” in Proc. 4th Int. Conf. Appl. Artif. Intell. Comput. (ICAAIC), 2025, pp. 198–204. [8] J. Kolli, D. M. Vamsi, and V. M. Manikandan, “Plant disease detection using convolutional neural network,” in Proc. IEEE Bombay Sect. Signature Conf. (IBSSC), 2021, pp. 1–6. [9] M. Sandler, A. Howard, M. Zhu, A. Zhmoginov, and L.-C. Chen, “MobileNetV2: Inverted residuals and linear bottlenecks,” in Proc. IEEE/CVF Conf. Comput. Vis. Pattern Recognit. (CVPR), 2018, pp. 4510–4520. [10] Y. Kim, R. G. Evans, and W. M. Iversen, “Remote sensing and control of an irrigation system using a distributed wireless sensor network,” IEEE Trans. Instrum. Meas., vol. 57, no. 7, pp. 1379–1387, Jul. 2008. [11] J. Gutiérrez, J. F. Villa-Medina, A. Nieto-Garibay, and M. Á. Porta-Gándara, “Automated irrigation system using a wireless sensor network and GPRS module,” IEEE Trans. Instrum. Meas., vol. 63, no. 1, pp. 166–176, Jan. 2014. [12] S. D. Khirade and A. B. Patil, “Plant disease detection using image processing,” in Proc. Int. Conf. Comput. Commun. Control Autom. (ICCUBEA), 2015, pp. 768–771. [13] V. Balafas, E. Karantoumanis, M. Louta, and N. Ploskas, “Machine learning and deep learning for plant disease classification and detection,” IEEE Access, vol. 11, pp. 114352–114377, 2023. [14] Z. Zhou, X. Chen, E. Li, L. Zeng, K. Luo, and J. Zhang, “Edge intelligence: Paving the last mile of artificial intelligence with edge computing,” Proc. IEEE, vol. 107, no. 8, pp. 1738–1762, Aug. 2019. [15] C. Suresh, S. Srikaran, T. Senthilkumar, V. Magizhan, and S. Santhosh, “An offline mobile application for plant-disease detection using MobileNetV3 and TensorFlow Lite,” in Proc. 7th Int. Conf. Comput. Vis. Bio Inspired Comput. (ICCVBIC), 2026, pp. 1463–1469. [16] Espressif Systems, “ESP32 series datasheet.” [Online]. Available: https://www.espressif.com/sites/default/files/documentation/esp32_datasheet_en.pdf

Copyright

Copyright © 2026 Tamilselvi K, Gowtham A, Dilip M, Blesson P. This is an open access article distributed under the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.