New Technology: AI-powered Automatic Detection and Stop for People and Vehicles from Drones in Flight (AeroFlex, a University of Tsukuba Startup)
AeroFlex, a University of Tsukuba startup, successfully experimented with a new technology where AI-equipped drones automatically detect and stop for people and vehicles during flight. This technology is crucial for ensuring safety in drone Level 3.5 flights and contributes to labor savings and reduced piloting difficulty.
📋 Article Processing Timeline
- 📰 Published: May 11, 2026 at 20:02
- 🔍 Collected: May 11, 2026 at 11:31
- 🤖 AI Analyzed: May 12, 2026 at 04:07 (16h 35m after Collected)
AeroFlex, a startup from the University of Tsukuba (Headquarters: Tsukuba City, Ibaraki Prefecture, Representative: Itsuki Horii), in collaboration with B&PLUS Inc., which conducts R&D on wireless power supply and charging devices for mobility, conducted an experiment on a technology that uses AI analysis of video captured by drones in flight to detect people and vehicles when crossing roads and automatically stop the flight. This new technology is indispensable for ensuring ground safety by confirming pedestrians and other objects with cameras, which is a condition for drone Level 3.5 flights.
AI-equipped drone developed
AI system for aerial safety confirmation during Level 3.5 flight
Conventional Level 3 drone flights required "placement of assistants or public notification with signs" as a measure to prohibit third parties from entering the flight path. The new Level 3.5 flight allows this to be replaced by "confirmation of the presence of pedestrians, etc., by an onboard camera." This makes it easier than before to cross roads, railways, etc., in uninhabited areas.
Furthermore, if drones can autonomously analyze onboard camera footage and stop flying, it becomes possible to achieve labor savings and reduce piloting difficulty. Therefore, AeroFlex developed an AI system for drones that is mounted on the drone itself to "real-time AI analyze video from an onboard drone camera during flight and automatically control the flight," and conducted flight experiments by mounting it on its self-developed drone.
Flight experiment: Onboard camera confirmation and temporary stop
The drone used in this experiment is equipped with two onboard cameras and flies while capturing images directly below and in front. The AI analyzes the video in real-time, and when it recognizes people, vehicles, animals, etc., it automatically stops flying and waits in the air until the object moves out of the area.
In the experiment, we were able to actually confirm the series of steps where the AI system recognized a person and autonomously controlled the drone.
Drone equipped with two onboard cameras
Left: Image directly below, Right: Image in front (person detected by AI analysis)
Watch the experiment here ↓
Towards further extension of human eyes and will
Currently, human visual monitoring is required for drone flights, but this experiment confirmed the possibility of replacing human eyes with cameras and AI by combining AI technology. AeroFlex is further advancing the R&D and social implementation of next-generation mobility, aiming for labor saving and reduced manpower, by combining technologies such as wireless charging (with B&PLUS Inc.).
[About AeroFlex, a University of Tsukuba Startup]
It is a University of Tsukuba startup that develops, manufactures, and sells drones, robots, and unmanned systems. Looking beyond next-generation mobility, we are engaged in R&D of spacecraft and rocket engine systems, and mobile sensing technology that extends human eyes and will.
We are also actively engaged in joint research and development with research institutions and other companies (industry-academia collaboration). We also accept custom development of robots and drones.
[Contact Information]
TEL: 080-6974-1239
E-mail: contact@aeroflex.tech
Inquiry Form: https://www.aeroflexllc.com/contact-5
AI-equipped drone developed
AI system for aerial safety confirmation during Level 3.5 flight
Conventional Level 3 drone flights required "placement of assistants or public notification with signs" as a measure to prohibit third parties from entering the flight path. The new Level 3.5 flight allows this to be replaced by "confirmation of the presence of pedestrians, etc., by an onboard camera." This makes it easier than before to cross roads, railways, etc., in uninhabited areas.
Furthermore, if drones can autonomously analyze onboard camera footage and stop flying, it becomes possible to achieve labor savings and reduce piloting difficulty. Therefore, AeroFlex developed an AI system for drones that is mounted on the drone itself to "real-time AI analyze video from an onboard drone camera during flight and automatically control the flight," and conducted flight experiments by mounting it on its self-developed drone.
Flight experiment: Onboard camera confirmation and temporary stop
The drone used in this experiment is equipped with two onboard cameras and flies while capturing images directly below and in front. The AI analyzes the video in real-time, and when it recognizes people, vehicles, animals, etc., it automatically stops flying and waits in the air until the object moves out of the area.
In the experiment, we were able to actually confirm the series of steps where the AI system recognized a person and autonomously controlled the drone.
Drone equipped with two onboard cameras
Left: Image directly below, Right: Image in front (person detected by AI analysis)
Watch the experiment here ↓
Towards further extension of human eyes and will
Currently, human visual monitoring is required for drone flights, but this experiment confirmed the possibility of replacing human eyes with cameras and AI by combining AI technology. AeroFlex is further advancing the R&D and social implementation of next-generation mobility, aiming for labor saving and reduced manpower, by combining technologies such as wireless charging (with B&PLUS Inc.).
[About AeroFlex, a University of Tsukuba Startup]
It is a University of Tsukuba startup that develops, manufactures, and sells drones, robots, and unmanned systems. Looking beyond next-generation mobility, we are engaged in R&D of spacecraft and rocket engine systems, and mobile sensing technology that extends human eyes and will.
We are also actively engaged in joint research and development with research institutions and other companies (industry-academia collaboration). We also accept custom development of robots and drones.
[Contact Information]
TEL: 080-6974-1239
E-mail: contact@aeroflex.tech
Inquiry Form: https://www.aeroflexllc.com/contact-5