About

Data Collection

At present, AutoMine supports two major autonomous driving tasks. More tasks will be taken into account in the future.

Collection platforms

We utilizing three acquisition platforms (Touareg SUV, a Tonly TLD65 mining wide-body transport truck and a Komatsu 930-4E mining truck) at five strip mining sites in Inner Mongolia and Shanxi province of China. Each collection vehicle contains at least one front lidar, an inertial navigation system, and two monocular cameras. These vehicles’ appearance and sensor setup for the wide-body truck can be seen as follow:

Sensors

Two FLIR industrial cameras were mounted on the top of three acquisition platforms with 55Hz capture frequency, 1/1.8” anamorphic format, 2048×1536 resolution, 70 degrees Field of View (FOV). The Velodyne HDL-32E has 32 beams, 20Hz capture frequency, 360° horizontal FOV (only 180° valid), -30° ∼ +10° vertical scan scope, 70m range, ±2cm accuracy, up to 1.4M points per second and the Ouster-II-64 lidar has 64 beams with -7.9° ∼ +7.9° vertical angles, 20Hz, 150m range, up to 1.3M points per second. The inertial navigation system updates it’s frequency in 10Hz. Table shows models of other core sensors.

 The SUVThe Wide-Body TruckThe Mining Car
Main LidarOuster OS-2-64 * 1Velodyne HDL-32E * 1Velodyne HDL-32E * 1
CameraBFS-U3-31S4C-C *2BFS-U3-31S4C-C *2BFS-U3-31S4C-C *2
Inertial Navigation SystemDAISCH * 1DAISCH * 1DAISCH * 1
Solid LaserLivox MID-40Livox MID-40
Edge LidarVelodyne VLP-16 * 3Velodyne VLP-16 * 3
Blinding lidarRS-Bpearl * 3RS-Bpearl * 3
RadarARS408-21 * 1ARS408-21 * 3

Each student has access to the best possible learning technologies, as well having guest classes from industry professionals. We believe this better prepares your child for their careers or own businesses.

Sensor synchronization

The classroom environment at Stratford allows your childs’ educators to give them th

We utilizing three acquisition platforms (Touareg SUV, a Tonly TLD65 mining wide-body transport truck and a Komatsu 930-4E mining truck) at five strip mining sites in Inner Mongolia and Shanxi province of China. Each collection vehicle contains at least one front lidar, an inertial navigation system, and two monocular cameras. These vehicles’ appearance and sensor setup for the wide-body truck can be seen as follow:

Annotation

For localization, the data includes position information (longitude, latitude and altitude) provided by the GPS and kinematic information (the speed and acceleration) supported by the inertial measure unit (IMU) of the collection vehicles. Perceptual annotation data consists of targets’ 2D bounding boxes on image coordination, 3D properties with location (x,y,z), dimension (h,w,l), and rotation (roll,pitch,yaw), classification including trucks, tractors, trailers, wide-body trucks, mining trucks, excavators, pushdozers, civilian vehicles and pedestrians.

Contact Us

We are Automine team, free to support with comments, feedback.