D.Hive make autonomous platform for drones provide two functions.
- Fully avoidance
- Automated Path Planning

Why this need.
1. Industrial drones are increasing.
According to FAA aerospace forecast, the number of industrial drones is increasing from 277,000 in 2018 to 1.29 million in 2023. Also, market size is expected $12.7B in 2023 worldwide.
2. But drones are blind.
Despite the expectation for drones, blindness is holding back drones from being used throughout the industry. Because most of drones rely on GPS and IMU, those sensors cannot tell drones exactly where they are and what’s around it.
3. Rule of industrial drones.
Therefore, that limitation makes implicit rules. Each drone needs two experts. One looks at the environment around the drone and notify. The other monitors the drone’s view and directly control drone.
4. Low scalability
As the result, one drone requires labor cost of two experts annually. That expert makes entry barrier of training time and cost.

Howeve...
D.Hive make autonomous platform for drones provide two functions.
- Fully avoidance
- Automated Path Planning

Why this need.
1. Industrial drones are increasing.
According to FAA aerospace forecast, the number of industrial drones is increasing from 277,000 in 2018 to 1.29 million in 2023. Also, market size is expected $12.7B in 2023 worldwide.
2. But drones are blind.
Despite the expectation for drones, blindness is holding back drones from being used throughout the industry. Because most of drones rely on GPS and IMU, those sensors cannot tell drones exactly where they are and what’s around it.
3. Rule of industrial drones.
Therefore, that limitation makes implicit rules. Each drone needs two experts. One looks at the environment around the drone and notify. The other monitors the drone’s view and directly control drone.
4. Low scalability
As the result, one drone requires labor cost of two experts annually. That expert makes entry barrier of training time and cost.

However, our solution can solve this.
Our solution name is Advanced Intelligence for Universal Robotics, AIUR for short.
We provide two functions.
- Fully Avoidance
- Automated Path Planning
AIUR can be the eyes of drones like humans do. But little difference between human is that the eyes also rest on the back of the drones. So, drones can detect and avoid obstacles in every direction.
In detail, our solution is based on fisheye RGB camera (like T265 module of Intel). With 4 ~ 6 cameras (In case of T265 is 2 ~ 3ea), we can cover every direction. And AIUR can detect obstacle with motion detection and depth estimation model developed with deep learning. After that, moving obstacles problem is solved by avoidance model and fixed obstacle problem is solved by navigation model. Now our system is working on Jetson Tx2 device and can be compatible with other Linux based hardware. To consider compatible of drones, our system is compatible with commercial products like DJI and Pixhawk flight computer with MAVLink protocol. So, it can be applied most of industrial drones.

Right now, we start in agriculture sides. For crop dusting automation, we are talking with NH bank (for drone application) and agriculture drone manufactures.
Next target is infrastructure inspection for cracking monitoring inside to outside. In case of inside, construction company and outside, transmission tower and rail road can be applied with automation for detecting crack frequently and accurately with low cost.
More information

Employees

Jaeyeong Gim
Admin
Jaeyeong Gim CEO Growing white drawing paper, which can paint anything(but not everything).