Mandi Cast: AI-Based Crop Price Forecasting and Sell Advisory System

Author

Mrs. M. Manimegalai, Sanjay B,Siddarth J,Udayabhaarathi K

Abstract

 Mandi Cast is an AI-powered web application developed to forecast the daily wholesale (mandi) prices of major crops and to advise farmers whether to sell immediately or wait for a better price. In conventional agricultural marketing, farmers learn the price only on the day they reach the mandi, with no indication of the expected trend, and are therefore often forced to sell at low prices to intermediaries.

The proposed system collects historical variety-wise daily prices from the Government of India open data platform (Agmarknet / data.gov.in) and merges them with district-level weather data from OpenWeatherMap. Seven crops across five mandis are modelled using two complementary methods: a seasonal ARIMA model, SARIMA(1,1,1)(1,1,1,7), and an XGBoost regressor built on lag prices, rolling averages, calendar features, rainfall and temperature. For each crop–market pair both models are evaluated on a 30-day hold-out using RMSE, MAE and MAPE, and the model with the lower MAPE is saved, giving 35 deployed models.

A rule-based recommendation engine converts the seven-day forecast into SELL NOW, WAIT or STABLE advice with a confidence score. The system is delivered through a FastAPI backend and a React dashboard offering OTP-based login, an animated market carousel, price-trend sparklines and SMS price alerts. A preliminary evaluation on synthetic demo data gave a mean MAPE of 3.29% for the selected models. Mandi Cast thus converts raw price records into a simple, timely decision for farmers. Keywords: Crop Price Forecasting, SARIMA, XGBoost, Mandi Prices, Sell Advisory, Agricultural Decision Support, FastAPI, React.

Keywords

AI-based crop forecasting, Predictive agricultural analytics, • Machine learning price prediction, Mandi price forecasting, Crop price prediction

Full Text:

Download Paper PDF

References

[1] G. E. P. Box, G. M. Jenkins, G. C. Reinsel and G. M. Ljung, Time Series Analysis: Forecasting and Control, 5th ed. Hoboken, NJ, USA: Wiley, 2015.

[2] H. Adanacioglu and M. Yercan, "An analysis of tomato prices at wholesale level in Turkey: An application of SARIMA model," Custos e @gronegócio on line, vol. 8, no. 4, pp. 52–75, 2012.

[3] T. Chen and C. Guestrin, "XGBoost: A scalable tree boosting system," in Proc. 22nd ACM SIGKDD Int. Conf. Knowledge Discovery and Data Mining, San Francisco, CA, USA, 2016, pp. 785–794, doi: 10.1145/2939672.2939785.

[4] G. K. Jha and K. Sinha, "Agricultural price forecasting using neural network model: An innovative information delivery system," Agricultural Economics Research Review, vol. 26, no. 2, pp. 229–239, 2013.

[5] O. Ahumada and J. R. Villalobos, "Application of planning models in the agri-food supply chain: A review," European Journal of Operational Research, vol. 196, no. 1, pp. 1–20, 2009.

[6] R. J. Hyndman and A. B. Koehler, "Another look at measures of forecast accuracy," International Journal of Forecasting, vol. 22, no. 4, pp. 679–688, 2006, doi: 10.1016/j.ijforecast.2006.03.001.

[7] Department of Agriculture and Farmers Welfare, Government of India, "Variety-wise Daily Market Prices Data of Commodity," Open Government Data (OGD) Platform India, data.gov.in.

Share your valuable work from Social Media Buttons