Hyderabad: When are people more likely to encounter a snake? The Friends of Snakes Society (FOS) is using weather data and rescue records to answer the question, developing an hourly forecast of snake encounters. The predictions have already covered 72 locations across Hyderabad and its neighbouring districts.The machine-learning model, developed using 35,234 snake-rescue dispatches recorded across Telangana in 2024 and 2025, was presented on Oct 4 at First World Congress on Snakes in Kandy, Sri Lanka. The congress, held from Oct 1-4, also featured FOS presentations on urban snake ecology and its rescue network.Researchers matched rescue dispatches with hourly weather conditions and sunlight data to identify factors linked to snake encounters. The model was tested against records not used during training and subsequently against new rescue data from 2026.The model estimates the likely intensity of reported snake encounters for each hour. FOS said the forecasts could help residents take precautions and rescuers prepare for periods of higher demand. Forecasts for the next 24 hours are updated up to four times a day, while longer forecasts extend to 15 days.“We want people to know when they should be especially watchful around their homes, farms and workplaces,” said Avinash C Visvanathan, FOS general secretary and lead researcher.Experimental district-level forecasts for other parts of India are also being run on FOS’s Snakebite Insight platform. Over the next year, the society plans to train and evaluate the model using data from other regions, where climate, snake species and reporting patterns vary.Visvanathan said FOS would also explore whether such forecasts could help health services plan anti-snake venom availability. However, the current model predicts reported snake encounters, not snakebites, and would need to be assessed against actual bite data before being used for healthcare planning.Another FOS presentation analysed 55,467 snake rescues in Hyderabad between 2013 and 2022, examining seasonal, hourly and geographical patterns. A third presentation detailed its round-the-clock helpline, mobile app and rescuer network, which assigns calls to nearby rescuers and helps identify duplicate reports.
Source link
