In a meadow in Switzerland, a drone hovered gracefully, carrying what looked like an oversized cotton swab attached to a four-meter cord. But this wasn’t a simple flight — it was a biological expedition in the air. As the swab brushed through the grasslands and treetops, it gathered microscopic traces of life — fragments of skin and fur, pollen grains, fungal spores, and even minute droplets of blood or saliva. Later, scientists used a pocket-sized DNA sequencer to decode the material, revealing dozens of species, some endangered and others invasive, many of which had never been seen by the human eye. This is how machines are now helping humanity detect the invisible life that has always surrounded us.
Researchers from the Swiss Federal Institute for Forest, Snow and Landscape Research are pioneering new ways to monitor biodiversity, mapping insect populations across ecosystems. They represent part of a growing scientific movement using robotics, machine learning, and genetic sequencing to track the natural world with precision never before possible. “We know so much more about what’s happening,” says ecologist Camilla Alboui of ETH Zurich, “even if much still escapes us.”
These tools mark a new era of machine-assisted taxonomy — a field that once depended solely on human eyes and endless hours of manual classification. Now, autonomous drones, eDNA collection systems, and AI algorithms can scan forests, oceans, and skies to detect organisms from the tiniest microbes to complex mammals. This digital revolution allows biologists to catalog the Earth’s estimated 8 million species, of which only 2.3 million are officially described. The rest remain what scientists call “dark taxa” — unnamed, unstudied, and at risk of disappearing before they are even known.
For example, entomologists estimate that four out of five insect species have never been documented. Despite being the planet’s most numerous and ecologically vital organisms, insects remain largely a mystery. Their hidden diversity holds immense potential — from natural compounds that may lead to new medicines to behavioral patterns that could inspire sustainable agricultural innovations. Yet, until recently, such research was painfully slow.
Today, AI-powered image recognition and bioacoustic systems can detect animal species by sound or wing patterns. 3D scanning and high-throughput sequencing enable researchers to study structures invisible to the naked eye and decode DNA of organisms that can’t be cultured in labs. In Panama, an AI-powered camera setup identified over 2,000 species of moths in just one week — half of them previously unknown to science. Similarly, in Colombia, drones combined with machine learning helped researchers describe tens of thousands of species, including 200 new discoveries.
However, even this technological progress has limits. Artificial intelligence only recognizes what’s already in its training data. That means many unrecorded species remain invisible to machines, especially those from the Global South. The vast majority of biodiversity databases are dominated by species from North America and Europe — a bias that reduces AI’s accuracy when applied elsewhere. As computer scientist David Rolnick from McGill University notes, “AI lacks intuition and creativity — qualities only human taxonomists can bring.”
Despite the growing power of automation, the human element remains irreplaceable. The world’s museums hold billions of specimens, but only a fraction are digitized, and less than 0.2% have genomic data available. Many collections are deteriorating, threatening to erase irreplaceable biological records. Meanwhile, climate change accelerates species loss, disrupting ecosystems and pushing biodiversity into collapse. As entomologist Valerio Caruso from the University of Padua warns, “Once you lose one species, others soon follow — eventually, everything falls apart.”
In the end, machines are helping scientists race against time. They offer the speed and precision humanity needs to document life before it vanishes. But the mission to understand the planet’s living archive still depends on human curiosity and commitment. Each data point, each genetic sequence, brings us closer to preserving the world’s biological memory — before the countdown reaches zero.
