RootSense Multimodal Crop Disease Diagnosis with Soil Weather Fusion and Conversational Recommendations

DOI

10.2991/978-94-6239-678-4_43How to use a DOI?

Keywords

Multimodal Crop Disease Diagnosis; Ensemble Learning; Soil-Weather Data Fusion; Convolutional Neural Networks

Abstract

RootSense is an artificial intelligence-powered chatbot designed to assist farmers and farming in India by automating crop disease diagnosis and consulting services. Using photographs of crops uploaded by farmers, RootSense utilises hyperlocal environmental variables such as soil qualities, current weather, and location-specific agronomic data to provide context-aware and farm-specific recommendations, in contrast to existing methods of crop diagnosis using only computer vision techniques. In comparison, RootSense will extract data from the hyperlocal environment of each photograph uploaded by the farmer and combine it with the physical properties of the soil (i.e., moisture, pH, and nutrient levels) to create diagnostic suggestions to farmers that are not generalised or misdiagnosed for that specific farm. The architecture for RootSense includes five main components: (1) the requirements analysis for farmers, (2) the collection of multimodal data from the PlantVillage and NBSS and LUP databases, (3) the classification of crop diseases using CNNs, (4) the use of ensemble fusion to infer the most likely cause, and (5) the deployment of the service using the Node.js, React Native, and Cloud APIs platforms. RootSense will provide geographically-based (geo-location) real-time advice on crop management, pest control, and irrigation to help farmers in different agro climatic regions of India. Compared to the current benchmarks of 78–80% for image-only methods of diagnosis, the experimental results indicate that the RootSense service will produce diagnostic results of approximately 95–96%. Additionally, the diagnostic cause attribution accuracy will improve from over 50% to over 90%, with reduced false alarms.

Copyright

© 2026 The Author(s)

Open Access

Open Access This chapter is licensed under the terms of the Creative Commons Attribution-NonCommercial 4.0 International License (http://creativecommons.org/licenses/by-nc/4.0/), which permits any noncommercial use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license and indicate if changes were made.

Download article (PDF)

TY  - CONF
AU  - T. Lakshmana Kumar
AU  - R. Yasir Abdullah
AU  - M. Vijaykumar
AU  - T. Vinothkumar
AU  - R. Ranjith Bharathi
AU  - C. Hari Hara Kumar
PY  - 2026
DA  - 2026/05/28
TI  - RootSense Multimodal Crop Disease Diagnosis with Soil Weather Fusion and Conversational Recommendations
BT  - Proceedings of the 2nd International Conference on Recent Advancement and Modernization in Sustainable Intelligent Technologies & Applications (RAMSITA-2026)
PB  - Atlantis Press
SP  - 540
EP  - 553
SN  - 1951-6851
UR  - https://doi.org/10.2991/978-94-6239-678-4_43
DO  - 10.2991/978-94-6239-678-4_43
ID  - Kumar2026
ER  -