KrushiDoot: A Smart Platform Guiding Farmers in Crop Cultivation and Soil Management

Abstract

Crop analysis and prediction is a rapidly growing field which is vital in optimizing agricultural practices. Crop recommendation is pivotal in agriculture, empowering farmers to make informed decisions about the most suitable crops for their land and climate conditions. Traditionally, this process heavily relied on expert knowledge, which proved time-consuming and labor-intensive. Moreover, considering the projected global population of 9.7 billion by 2050, the need to produce more food sustainably becomes imperative. Machine learning techniques can play a crucial role in effectively automating crop recom- emendations and detecting pests and diseases to enable farmers to optimize their yield from the land while simultaneously maintaining soil fertility and replenishing essential nutrients. This paper analyses the performance of crop recommendation across seven distinct machine-learning algorithms. The proposed system leverages various features, including soil composition and climate data, to accurately predict the most suitable crops for specific locations. This system has the potential to revolutionize crop recommendation, benefiting farmers of all scales by enhancing crop yields, sustainability, and overall profitability. Through ex- tensive evaluation of a comprehensive historical data set, we have achieved near-perfect accuracy by training and testing models of machine learning algorithms with various configurations. We demonstrate accuracy consistently over 95% across all models, with the highest achieved accuracy reaching 99.5%.

Introduction

Agriculture remains critical for global food security but faces growing challenges from climate change, soil degradation, pests, and resource limitations. To address these issues, Artificial Intelligence (AI) and Machine Learning (ML) have been increasingly integrated into agricultural practices, enabling predictive, adaptive, and data-driven decision-making. Classical ML algorithms (Logistic Regression, Decision Trees, Naïve Bayes, Support Vector Machines) and ensemble models (Random Forests, Gradient Boosting) have been applied to crop classification, yield forecasting, and plant disease detection, while deep learning architectures (CNNs, RNNs, LSTMs) further improve pattern recognition and temporal prediction. Frameworks like TensorFlow and PyTorch, coupled with optimizers (Adam, Adagrad) and activation functions (ReLU, Tanh), allow scalable processing of multi-dimensional agricultural data.

Climate variability and environmental challenges have necessitated adaptive, ML-driven forecasting for weather anomalies, crop failure risk, and optimal cultivation schedules. In response, KrushiDoot has been developed as a web-based, modular AI platform that integrates multiple ML and generative AI components for holistic crop management. Unlike sensor-heavy systems, KrushiDoot emphasizes accessibility, explainability, and decision intelligence. Its modules include:

• Crop Suitability Assessment: Ensemble models using Random Forest, Gradient Boosting, and Linear Regression.

• Yield Forecasting: LSTM-based temporal modeling incorporating weather, satellite-derived vegetation indices, and historical yield data.

• Plant Disease Detection: CNN-based ResNet50 transfer learning on PlantDoc dataset with Amazon API integration for treatment recommendations.

• Flood Risk Prediction: Hybrid ML models.

• Dynamic Crop Timeline: Generative AI (Groq API, Llama 3 70B) provides contextual, natural language advice across growth stages.

KrushiDoot’s architecture comprises five layers: user interface, security, application, ML data processing, and hybrid data storage (PostgreSQL + MongoDB). Standardized data preprocessing, augmentation, feature engineering, and stratified sampling ensure model reliability. The system achieves high accuracy (disease detection >95%, crop suitability R² = 0.87, yield forecasting MAPE = 11.8%) and maintains >99% uptime with low response times. Evaluation incorporates precision, recall, F1-score, and expert validation, with continuous monitoring and retraining to adapt to evolving conditions.

The literature review highlights the evolution from classical ML to deep learning and generative AI in agriculture, emphasizing the importance of model integration, scalability, and interpretability. KrushiDoot addresses gaps in accessibility, modularity, and user-centered design by combining traditional ML precision with LLM-based contextual intelligence.

Future Directions:

1. Conduct farmer surveys to evaluate economic impact.

2. Develop a mobile application for broader accessibility.

3. Collect regional data for model adaptability.

4. Use larger and more diverse datasets to improve predictions.

5. Assess economic and environmental benefits.

6. Integrate IoT sensors for real-time data and improved decision-making.

Conclusion

This paper presents how AI-based platform KrushiDoot can improve agricultural practices by using machine learning and deep learning models. Models like LSTM for yield prediction, ensemble methods for crop recommendation, and ResNet50 for disease detection help farmers get accurate and timely suggestions. By combining real-time weather data and crop management insights, the system helps improve productivity, resource use, and decision-making. The suggested method can be used in many places and on many types of farms because it can be changed to fit different needs. It can also help farmers, governments, and agri-businesses make better choices and come up with smart ways to farm. Generative and explainable AI models could be added to KrushiDoot in the future to make it easier to understand and get better advice. In general, this research helps to create a farming ecosystem that is more sustainable, efficient, and smart.

References

[1] A. L. Samuel, “Some studies in machine learning using the game of checkers,” IBM Journal of Research and Development, vol. 3, no. 3, pp. 210–229, 1959. [2] H. Wang, Z. Lei, X. Zhang, B. Zhou, and J. Peng, “Machine learning basics,” Deep learning, pp. 98–164, 2016. [3] G. Bonaccorso, Machine learning algorithms. Packt Publishing Ltd, 2017. [4] A. Calzadilla, T. Zhu, K. Rehdanz, R. S. Tol, and C. Ringler, “Climate change and agriculture: Impacts and adaptation options in south africa,” Water Resources and Economics, vol. 5, pp. 24–48, 2014. [5] T. Partap, “Hill agriculture: challenges and opportunities,” Indian Jour- nal of Agricultural Economics, vol. 66, no. 902-2016-67891, 2011. [6] T. O. Ayodele, “Types of machine learning algorithms,” New advances in machine learning, vol. 3, pp. 19–48, 2010. [7] D. W. Hosmer Jr, S. Lemeshow, and R. X. Sturdivant, Applied logistic regression. John Wiley & Sons, 2013, vol. 398. [8] R. E. Wright, “Logistic regression.” 1995. [9] B. Charbuty and A. Abdulazeez, “Classification based on decision tree algorithm for machine learning,” Journal of Applied Science and Technology Trends, vol. 2, no. 01, pp. 20–28, 2021. [10] M. R. Segal, “Machine learning benchmarks and random forest regres- sion,” 2004. [11] K. Taunk, S. De, S. Verma, and A. Swetapadma, “A brief review of nearest neighbor algorithm for learning and classification,” in 2019 International Conference on Intelligent Computing and Control Systems (ICCS). IEEE, 2019, pp. 1255–1260. [12] S. Chen, G. I. Webb, L. Liu, and X. Ma, “A novel selective naïve bayes algorithm,” Knowledge-Based Systems, vol. 192, p. 105361, 2020. [13] M. A. Hearst, S. T. Dumais, E. Osuna, J. Platt, and B. Scholkopf, “Sup- port vector machines,” IEEE Intelligent Systems and their applications, vol. 13, no. 4, pp. 18–28, 1998. [14] I. Steinwart and A. Christmann, Support vector machines. Springer Science & Business Media, 2008. [15] K. Gurney, An introduction to neural networks. CRC press, 1997. [16] S. Albawi, T. A. Mohammed, and S. Al-Zawi, “Understanding of a convolutional neural network,” in 2017 International Conference on Engineering and Technology (ICET), 2017, pp. 1–6. [17] S. Sharma, S. Sharma, and A. Athaiya, “Activation functions in neural networks,” Towards Data Sci, vol. 6, no. 12, pp. 310–316, 2017. [18] J. Dean and S. Ghemawat, “Mapreduce: Simplified data processing on large clusters,” in OSDI’04: Sixth Symposium on Operating System Design and Implementation, San Francisco, CA, 2004, pp. 137–150. [19] D. Dahiphale, R. Karve, A. V. Vasilakos, H. Liu, Z. Yu, A. Chhajer,J. Wang, and C. Wang, “An advanced mapreduce: Cloud mapreduce, enhancements and applications,” IEEE Transactions on Network and Service Management, vol. 11, no. 1, pp. 101–115, 2014. [20] R. Kiveris, S. Lattanzi, V. Mirrokni, V. Rastogi, and S. Vassilvitskii, “Connected components in mapreduce and beyond,” in SOCC 2014, 2014. [Online]. [21] D. Dahiphale, “Mapreduce for graphs processing: New big data algo- rithm for 2-edge connected components and future ideas,” IEEE Access, vol. 11, pp. 54 986–55 001, 2023. [22] A. Oikonomidis, C. Catal, and A. Kassahun, “Deep learning for crop yield prediction: a systematic literature review,” New Zealand Journal of Crop and Horticultural Science, vol. 51, no. 1, pp. 1–26, 2023. [23] P. A, S. Chakraborty, A. Kumar, and O. R. Pooniwala, “Intelligent crop recommendation system using machine learning,” in 2021 5th Inter- national Conference on Computing Methodologies and Communication (ICCMC), 2021, pp. 843–848. [24] Z. Doshi, S. Nadkarni, R. Agrawal, and N. Shah, “Agroconsultant: Intel- ligent crop recommendation system using machine learning algorithms,” in 2018 Fourth International Conference on Computing Communication Control and Automation (ICCUBEA), 2018, pp. 1–6. [25] S. M. PANDE, P. K. RAMESH, A. ANMOL, B. R. AISHWARYA, K. ROHILLA, and K. SHAURYA, “Crop recommender system using machine learning approach,” in 2021 5th International Conference on Computing Methodologies and Communication (ICCMC), 2021, pp. 1066–1071. [26] R. K. Rajak, A. Pawar, M. Pendke, P. Shinde, S. Rathod, and A. Devare, “Crop recommendation system to maximize crop yield using machine learning technique,” International Research Journal of Engineering and Technology, vol. 4, no. 12, pp. 950–953, 2017. [27] D. Reddy and M. R. Kumar, “Crop yield prediction using machine learning algorithm,” in 2021 5th International Conference on Intelligent Computing and Control Systems (ICICCS), 2021, pp. 1466–1470 [28] R. Ghadge, J. Kulkarni, P. More, S. Nene, and R. Priya, “Prediction of crop yield using machine learning,” Int. Res. J. Eng. Technol.(IRJET), vol. 5, 2018. [29] N. H. Kulkarni, G. N. Srinivasan, B. M. Sagar, and N. K. Cauvery, “Improving crop productivity through a crop recommendation system using ensembling technique,” in 2018 3rd International Conference on Computational Systems and Information Technology for Sustainable Solutions (CSITSS), 2018, pp. 114–119. [30] S. Pudumalar, E. Ramanujam, R. H. Rajashree, C. Kavya, T. Kiruthika, and J. Nisha, “Crop recommendation system for precision agriculture,” in 2016 Eighth International Conference on Advanced Computing (ICoAC), 2017, pp. 32–36. [31] K. G. Liakos, P. Busato, D. Moshou, S. Pearson, and D. Bochtis, “Machine learning in agriculture: A review,” Sensors, vol. 18, no. 8, p. 2674, 2018. [Online]. [32] M. Ayaz, M. Ammad-Uddin, Z. Sharif, A. Mansour, and E.-H. M. Ag- goune, “Internet-of-things (iot)-based smart agriculture: Toward making the fields talk,” IEEE Access, vol. 7, pp. 129 551–129 583, 2019. [33] V. Iglovikov, S. Mushinskiy, and V. Osin, “Satellite imagery feature de- tection using deep convolutional neural network: A kaggle competition,” arXiv preprint arXiv:1706.06169, 2017. [34] M. Abadi, A. Agarwal, P. Barham, E. Brevdo, Z. Chen, C. Citro, G. S. Corrado, A. Davis, J. Dean, M. Devin et al., “Tensorflow: Large-scale machine learning on heterogeneous distributed systems,” arXiv preprint arXiv:1603.04467, 2016. [35] S. Sharma, S. Sharma, and A. Athaiya, “Activation functions in neural networks,” Towards Data Sci, vol. 6, no. 12, pp. 310–316, 2017. [36] B. Ding, H. Qian, and J. Zhou, “Activation functions and their character- istics in deep neural networks,” in 2018 Chinese Control And Decision Conference (CCDC), 2018, pp. 1836–1841. [37] J. Brownlee, “What is the difference between a batch and an epoch in a neural network,” Machine Learning Mastery, vol. 20, 2018. [38] J. T. Barron, “A general and adaptive robust loss function,” in Pro- ceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, 2019, pp. 4331–4339. [39] D. P. Kingma and J. Ba, “Adam: A method for stochastic optimization,” arXiv preprint arXiv:1412.6980, 2014. [40] A. Lydia and S. Francis, “Adagrad—an optimizer for stochastic gradient descent,” Int. J. Inf. Comput. Sci, vol. 6, no. 5, pp. 566–568, 2019. [41] A. M. Taqi, A. Awad, F. Al-Azzo, and M. Milanova, “The impact of multi-optimizers and data augmentation on tensorflow convolutional neural network performance,” in 2018 IEEE Conference on Multimedia Information Processing and Retrieval (MIPR), 2018, pp. 140–145. [42] P. Gulati, A. Sharma, and M. Gupta, “Theoretical study of decision tree algorithms to identify pivotal factors for performance improvement: A review,” Int. J. Comput. Appl, vol. 141, no. 14, pp. 19–25, 2016. [43] J. Davis and M. Goadrich, “The relationship between precision-recall and roc curves,” in Proceedings of the 23rd international conference on Machine learning, 2006, pp. 233–240. [44] T. B. Pathak, M. L. Maskey, J. A. Dahlberg, F. Kearns, K. M. Bali, and D. Zaccaria, “Climate change trends and impacts on california agriculture: a detailed review,” Agronomy, vol. 8, no. 3, p. 25, 2018. [45] N. Mancosu, R. L. Snyder, G. Kyriakakis, and D. Spano, “Water scarcity and future challenges for food production,” Water, vol. 7, no. 3, pp. 975– 992, 2015. [46] PlantDoc Dataset Documentation (2023). Dataset Summary for Visual Plant Disease Detection. DatasetNinja repository [47] Sharma et al. (2024). \"Prediction of best suitable crop using machine learning technique.\" International Journal of Intelligent Systems and Applications in Engineering, 12(23s), 2743 [48] IJISAE (2024). \"Prediction of best suitable crop using machine learning technique: Ensemble stacking methodology\" [49] Yadav et al. (2025). \"Time series forecasting of crop yield under climate variability using IoT data and LSTM networks.\" International Journal of Environmental Sciences [50] Peeriga et al. (2024). \"Real-time rain prediction in agriculture using AI and IoT: A bi-directional LSTM approach.\" Engineering, Technology & Applied Science Research, 14(4), 15805-15812 [51] ADaSci (2024). \"Implementing rapid LLM inferencing using Groq.\" Retrieved from https://adasci.org [52] DataCamp (2024). \"Groq LPU inference engine tutorial.\" LLamaIndex framework documentation [53] GeeksforGeeks (2025). \"Beginner\'s guide to Groq API with Llama 3.\" NLP and AI tutorials [54] OpenWeather (2025). \"Weather data for agriculture -- Forecasts & crop insights.\" OpenWeather Agriculture Solutions [55] OpenWeather API Documentation (2024). Historical weather data and forecasting services [56] Singh et al. (2020). \"PlantDoc: A dataset for visual plant disease detection.\" ACM India Joint International Conference on Data Science and Management of Data (CoDS-COMAD 2020) [57] He et al. (2015). \"Deep Residual Learning for Image Recognition.\" arXiv:1512.03385

Copyright

Copyright © 2025 Prof. Sandip Shinde, Parth Kedari, Atharva Khaire, Shaunak Karvir, Omkar Kumbhar. 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.