Although neuromorphic computing was first proposed by Carver Mead in the late 1980s, this area of computer design is still evolving. (Image credit: Anton Petrus/Getty Images)
Researchers are creating artificial intelligence (AI) on a chip the size of a grain of rice that can mimic the functioning of the human brain and plan to use it in tiny drones.
While AI can automate routine tasks, it requires significant resources and consumes a lot of energy to operate. Drones also require energy to move, navigate, sense, stabilize, and communicate.
Larger drones can more easily handle AI's power requirements using a motor, but smaller drones run on batteries – meaning AI's power requirements can reduce flight time from 45 minutes to just four.
That could be a temporary problem, however. Suyin Yi and his team at the University of Texas have received funding from the 2025 Air Force Office of Scientific Research Young Investigator Program (part of the Air Force Office of Scientific Research) to develop energy-efficient AI for drones. They aim to create a chip the size of a grain of rice with a variety of AI functions, including autonomous piloting and object recognition, within three years.
Miniature drones with artificial intelligence
To develop a more energy-efficient AI chip, scientists are proposing to use thin films of conductive polymer. This is a (still) under-explored aspect of neuromorphic computing; a computer system that mimics the structure of the brain to enable highly efficient information processing.
The researchers aim to replicate the learning and decision-making process of neurons, which will save energy because they are only activated when needed, similar to how the human brain uses different areas to perform different tasks.
Although neuromorphic computing was first proposed by Carver Mead in the late 1980s, this area of computer design is still evolving. In 2024, Intel unveiled its Hala Point neuromorphic computer, which runs on more than 1,000 new AI chips and performs calculations 50 times faster than traditional systems.
Meanwhile, the Joint Artificial Intelligence Center is developing both AI software and neuromorphic hardware. Their main focus is on creating systems that can share all sensory data between participants in a network of neuromorphic devices. This technology could greatly improve situational awareness, and its current applications include headsets and robotics.
Using technologies developed in this research, drones could become more intelligent by integrating systems made of conductive polymer materials that could function similarly to neurons in the brain.
If Yi’s research is successful, miniature drones could become significantly smarter. An AI system using neuromorphic computing could enable smaller, smarter automated drones to perform remote monitoring in confined spaces with significantly increased flight times.
TOPICS drones
Peter Ray Ellison
Peter is a qualified engineer and an experienced freelance journalist who specialises in science, technology and culture. He has written for a range of publications including the BBC, Computer Weekly, IT Pro, The Guardian and The Independent. Peter has been working in technology journalism for over ten years. He holds a degree in Computer Engineering from Sheffield Hall University
Sourse: www.livescience.com
