It can be seen from the spraying results that the effect of agricultural spraying drones and human operations is not significant, but the time-consuming is greatly reduced, and the utilization rate of water and pesticides is greatly improved. We can see that drone spray technology can replace part of the ground action method and achieve good results.

DJI T16, the second-generation plant protection drone

Fully automatic operation

It is very easy to operate. First, use a Phantom 4 RTK aerial survey drone to fly around the farmland to automatically take high-definition photos, and then transfer the data to a laptop for processing;

DJI Maps software will perform 3D modeling and AI route planning, and then insert the data card into the T16 drone, and they can automatically perform spraying operations according to the calculated route.

DJI drones can now perform fully automatic operations in irregular orchards, automatically avoiding obstacles such as buildings, ponds, and power lines. The entire process does not require human intervention.

During the entire operation, the drone pilot spends most of his time preparing the next round of pesticide spraying, rather than following the drone to operate, which greatly improves work efficiency.

DJI T16 is currently China’s first plant protection UAV that implements two AI technologies: 3D scene reconstruction and map semantic recognition. Working in the fruit tree scene, through aerial survey planning, after importing and processing DJI Terra, the job scene can be recognized. Fruit trees, buildings, telephone poles, rivers, open spaces and other details will not be missed.

After completing the job scene recognition and generating the 3D route, the T16 can perform automatic flight operations according to the planned route. Compared with the traditional uniform spraying mode of the entire operation area, the drone can save 30% of the use of pesticides and water.

In the orchard scene, there are various types of fruit trees, different heights, and complex terrain. The grass and dense fruit trees are easy to be confused and difficult to distinguish from a long-distance bird’s-eye view. T16 fruit tree mode adopts 3D point cloud slice and 3D feature fusion technology, which can correctly identify a variety of objects in complex scenes.

At present, this system has passed the training of more than 300,000 fruit tree pictures, and can complete 35Acre full-environment recognition in less than 1 minute. There are 20 types of fruit trees that can be recognized, including apples, citruses, mangoes, bananas, dragon fruits, etc. The accuracy of fruit tree recognition reached 95%.

In spraying operations, one pilot can make up to five T16 drones take off at the same time.

In most orchards, the planting of fruit trees is unevenly distributed, and there are certain differences in tree height due to differences in tree age. There are often obstacles such as buildings, telephone poles, and ponds in orchards, which brings great challenges to the work of drones. The emergence of the DJI T16 UAV “fruit tree model” is undoubtedly a revolution in the agricultural UAV market.

Orchard mode

Through more mature artificial intelligence technology, DJI has solved the needs of plant protection in orchards.

After completing the job scene recognition through AI, DJI Agras T16 understands the distribution of fruit trees and guarantees that Agras T16 is sprayed only where there are fruit trees, and it stops where there are no fruit trees. All this is performed automatically.

The DJI team also launched the “Tree Core Recognition” function, which uses AI to identify the position of the tree core in the orchard in just a few seconds, and can spray the center of the plant in a targeted manner.

The fully automatic way of working can make drone operators safer.

In the past, agricultural machinery operations in orchards could easily cause pesticide harm to people; with the help of artificial intelligence, drone operators can stay away from the work area for security monitoring.

The DJI T16 drone does not require manual operation when spraying pesticides. We can observe the operation of the equipment through the remote control.

The second-generation plant protection drone Agras T16

The new DJI T16 machine has reshaped the overall structure and adopts a modular design, bringing unprecedented high load and efficient spraying range.

“When we designed the T16, we first designed the wind field and then the other structures of the drone.” said Xie Tiandi, public relations director of DJI. “So the performance of this product is far superior to the previous generation. There is no other company on the market that has launched such a drone.”

The DJI T16 has a pesticide carrying capacity of up to 16 L. More importantly, in its six-rotor design, two auxiliary rotors are specially used to adjust the wind field. Directly below the T16, the airflow does not circulate like other drones. This design avoids touching unrelated areas during pesticide spraying and improves the spraying efficiency.

The researchers also found that using drones to spray pesticides can bring an unexpected benefit-the airflow brought by the drone’s rotor will lift the plant leaves, so that the sprayed liquid can reach deep areas such as the back of the plant leaves and the trunk. .

Our company hopes to continue to cooperate with the Canadian Agricultural Research Institute to extend drone technology to more types of crops, and promote the “precision agriculture ecosystem” to encourage farmers to get in touch with new technologies.

By Jenny Chen

July, 2020