R&D Platform-Chris Lab
CHRIS LAB (Innovative Human-Machine Intelligent System Laboratory) is located in Beijing Economic Development Zone, China. It is mainly engaged in the research and research of robot-related synchronous positioning and navigation algorithms, mapping technology, robotic systems, image monitoring, recognition, segmentation and other technologies. Provide a complete set of artificial intelligence solutions for household, commercial and industrial applications.
R & D team
In the R&D and engineering team, more than 85% of the staff with master's and doctoral qualifications have applied for more than 200 patents, and established long-term and stable technical cooperation with well-known universities at home and abroad, researching some technologies and product performance indicators in the field All have reached the international advanced level.
The basic research directions of the laboratory include synchronous mapping and positioning algorithms, robotic systems (perception, mapping, positioning, Multi-sensor fusion, control systems, motion control and path planning).
Leading SLAM Synchronous Positioning and Mapping Technology
SLAM technology is based on the lidar and camera modules carried by the robot, combined with core algorithms, the robot calculates its own position according to the sensor information, while constructing an environmental map, solving the problem of positioning and map construction when the mobile robot moves in an unknown location environment. And this real-time positioning and mapping technology are independent intellectual property rights, reaching the international advanced level and leading domestically. The application of this research in logistics robots, sweeping robots, industrial intelligent production and other fields has been fully verified. It realizes functions such as rapid establishment of maps and autonomous planning of travel paths in a brand-new environment, and can autonomously identify the road conditions and obstacles ahead. In a changing road environment, intelligent obstacle avoidance, autonomous route planning, and use of visual perception technology to build a three-dimensional dense map can complete the robot's spatial positioning within milliseconds, with positioning accuracy reaching the centimeter level.
Aiming at the working characteristics and working environment of mobile robots, a set of control systems for sweeping robots are designed. It adopts multi-sensor synchronization technology and multi-core and multi-threaded computing speed. The real-time kernel and task scheduling are optimized to perceive and respond to the environment in real time. Sending commands in a timely manner improves the efficient and flexible execution of the sweeping robot.
Positioning is the most basic link in the autonomous navigation of mobile robots, including sweeping robots. The high-precision and high-frequency two-dimensional and three-dimensional positioning technology developed by our company is a positioning technology based on structural features and uses adaptive feature extraction technology for complex scenes. , A positioning technology with high real-time positioning and high positioning accuracy is realized.
Build a Map
When the sweeping robot moves in an unknown environment, it performs real-time environment perception, real-time environment reconstruction, multi-viewpoint fusion and point cloud segmentation technologies. During the movement, it locates itself based on position estimation and sensor data, and at the same time builds incremental maps.
The sweeping robot integrates multi-source information such as lidar, camera, inertial components, ultrasound, wheel shorthand, and multi-sensor online calibration technology, which can eliminate system uncertainties, provide accurate observation results and intelligent data processing of comprehensive information, and make up for The shortcomings of lidar and vision cameras take advantage of the advantages of both.
The perception system of the mobile robot researched and developed by our company has the recognition and detection of specific objects, especially the recognition and tracking prediction of moving objects, and related technologies such as image semantic segmentation, so that the mobile robot can stably and accurately perceive the complex unknown environment information.
Motion Control and Path Planning
The sweeping robot can quickly perform global planning/local planning based on comprehensive information, and adopt the optimal running route. Under the rules of virtual traffic, it can perform path planning and motion control without any difference during the cleaning process, such as avoiding obstacles and automatically returning Full.