A 40-gram, palm-sized AI micro-drone has achieved what its developers describe as an engineering first by autonomously detecting, tracking and killing a flying insect in mid-air without direct human control. The breakthrough could mark an important step towards developing autonomous drones capable of targeting mosquitoes, which spread diseases such as malaria, dengue, Zika, chikungunya and yellow fever.According to a TOI report, the drone was developed by US startup Tornyol and successfully intercepted a flying moth during a controlled indoor demonstration. While the technology remains experimental, the company believes it could eventually offer a chemical-free approach to mosquito control.— alextoussss (@alextoussss) How the AI micro-drone achieved the autonomous mid-air insect killThe demonstration involved a tiny quadcopter weighing just 40 grams, roughly the size of a person's palm.According to TOI, after detecting a flying moth, the drone automatically tracked its movement, adjusted its flight path and intercepted the insect in mid-air without manual piloting.Tornyol says this is the first successful autonomous airborne insect interception by a micro-drone.However, the demonstration took place indoors using a hardware-in-the-loop system. External motion-capture cameras monitored the drone, while an offboard computer processed much of the flight data before sending commands back to the aircraft.The company says future versions are expected to perform all sensing, processing and navigation onboard.AI drone uses sonar instead of conventional camerasUnlike many autonomous drones that rely primarily on cameras, Tornyol's system uses ultrasonic sonar technology similar to the sensors used in vehicle parking systems.The drone emits ultrasonic waves and analyses the returning echoes to detect flying insects.It also combines miniature microphones with custom-built software to identify subtle Doppler shifts produced by insect wingbeats. The company says these unique wingbeat signatures allow the drone to detect insects that may be difficult to identify using conventional vision systems alone.Why mosquitoes are the ultimate targetAlthough the recent demonstration involved a moth, Tornyol says its long-term goal is to identify and eliminate mosquitoes.According to the company, the system is being developed to recognise mosquitoes by analysing their characteristic wingbeat frequencies. Future versions may even distinguish between different mosquito species and sexes.However, these capabilities are still under development and have not yet been independently validated through peer-reviewed scientific research.Researchers therefore describe the demonstration as an important proof of concept, rather than evidence that autonomous mosquito-hunting drones are ready for real-world deployment.Why eliminating mosquitoes mattersMosquitoes remain among the world's deadliest disease carriers. According to the World Health Organization (WHO), malaria alone caused an estimated 610,000 deaths in 2024, with children under five accounting for a significant proportion of fatalities.Mosquitoes also transmit:DengueZika virusChikungunyaYellow feverBecause many mosquito-control programmes rely heavily on chemical insecticides, scientists continue searching for more targeted and environmentally sustainable alternatives.Tornyol's vision for autonomous mosquito-hunting drone fleetsTornyol envisions networks of autonomous drones operating from compact charging stations installed around homes, parks and residential neighbourhoods.The drones would continuously patrol designated areas, detect mosquito wingbeats, intercept the insects and automatically return for recharging when required.The company claims the technology could reduce mosquito-control costs by up to 100 times compared with existing approaches and estimates that approximately 10 drones could protect one square kilometre.However, these figures are company projections and have not yet been demonstrated through large-scale field testing.Challenges before the technology becomes realityDespite the successful demonstration, several technical hurdles remain. Mosquitoes are significantly smaller, lighter and faster than moths, making them much more difficult to detect and intercept, particularly in outdoor environments where wind and weather can affect drone performance.Researchers must also overcome challenges involving:Battery lifeOnboard computing powerCollision avoidanceAccurate mosquito species identificationReliable operation in real-world conditions.Inputs from agencies