A flash of orange glides past on a forest trail. It settles briefly on a leaf before vanishing just as quickly. Was it a Plain Tiger? A Tawny Coster? Or something far rarer? Until now, answering that question often required the trained eye of a butterfly expert or hours spent poring over field guides. Today, all it takes is a photograph—and an app that works even where the internet doesn’t.
India is home to more than 1,450 butterfly species, making it one of the richest butterfly landscapes in the world. A new Android app, iButterfly Explorer, hopes to make this extraordinary diversity accessible to everyone. Developed by Dr Krushnamegh Kunte, Associate Professor at the National Centre for Biological Sciences (NCBS), and Kannan AS, founder of Nature’s Ark, a non-profit based in India, the app uses AI to identify Indian butterflies and provides detailed information on their natural history—all without requiring an internet connection.
“The app gives information about the vnatural history and biology of various butterfly species as well as information on legally protected species,” says Kunte, “The curated textual information that is embedded in the app is useful for beginners as well as experienced naturalists.” For Kannan, the goal was to make reliable butterfly identification available in the field. “This is the first Indian-made app to identify any Indian flora or fauna using modern AI algorithms,” he says. “We trained the model using a large database of expertly identified images from the Butterflies of India website. Once a photo is uploaded, users can also access information on caterpillars and larval host plants.”
But for Kunte, the app is as much about citizen science as technology. Observations by butterfly enthusiasts have already contributed to the discovery of species such as the Banded Tit (Hypolycaena narada) and, more recently, the Narrow-banded Wall (Chonala albistricta) from Arunachal Pradesh. The research is also helping conservation efforts. Long-term monitoring around Bengaluru has recorded increasing butterfly populations. “We submit our research findings to various state forest departments. They can use that information to identify areas where developmental activities may have minimum impacts on wildlife, and areas that host legally protected or sensitive butterfly species where such activities may be best avoided,” says Kunte, adding, “If people know which native plants support butterflies, they can create urban gardens that sustain them through difficult landscapes.”
As AI puts butterfly identification into more hands, the next significant discovery may not begin in a laboratory, but with someone stopping to photograph a butterfly on a forest trail—or even in their own backyard.