NextBrain™ AI Mushroom Cloud System

NextBrain™AI Mushroom
Cloud™ System

The NextBrain™ AI mushroom cloud system is a very powerful robot AI system from Steele. It adopts the research and development principle of robot biological anthropomorphism and has a robot brain, sensory nerves (multi-dimensional force sense, 3D vision, touch, sound), With autonomous motion trajectory planning, CSO algorithm (trajectory intelligent optimization), AI autonomous learning and training evolution, and automatic navigation, its product purpose is to make robot polishing truly intelligent and getting smarter.

Basic PARAMETERS

PRODUCT PARAMETERS

robot brain

Steel Robot Brain is a powerful robot control system that integrates perceptual neural networks and an excellent interactive experience with hardware from the bottom up. It is an important part of the NextBrain AI mushroom cloud system.

brain

perceptual neural network

force sense

1. The PolishX flexible force control system is a constant force compensation device system installed at the end of the robotic arm and based on pneumatic or electric principles. It can output axial, bidirectional controllable force according to work needs.
2. The iNS flexible force control system is a multi-dimensional force sensing system that can use fast trajectory generation technology and force position hybrid control technology to achieve:
a. Through iNS flexible force control technology and trajectory optimization technology, the processing speed of complex workpieces can be increased. At the same time, multi-dimensional constant force control and variable speed control can be implemented in the processing process to improve processing quality.
b. iNS flexible force control technology is combined with 3D vision technology. The robot learns through 3D vision and tactile perception action simulation (people hold the sensor to perform polishing actions), achieving high-precision automatic path planning without teaching.

Vision

Steele’s SeeV vision system forms 3D vision through dual-line cross-laser scanning or structured light cameras; it builds pixel-level 3D digital models and spatial relative position perception presentations, and works jointly with the polishing robot’s decision-making and control system.

touch

The STIL tactile system uses tactile sensors and AI algorithms to determine the touch of different materials and know what materials and textures they are, thereby providing material texture data in the AI mushroom cloud system and meeting the surface requirements of different materials in the polishing scene. Realize instant formulation of process recipes.

sound sense

By monitoring the sound vibration of the equipment, it is possible to control the noise in the working environment, analyze the operating status of the mechanical equipment, and achieve measurement and diagnosis of polishing quality. Based on the acoustic algorithm, the process polishing sound data is provided in the AI Mushroom Cloud system, and the process formula can be formulated in real time for different materials and different surface requirements in the polishing scene.

CSO Chicken Swarm Algorithm

Steel CSO chicken flock optimization algorithm is a new biological optimization algorithm. CSO simulates the hierarchical order of chicken flocks and the behavior of chicken flocks (including roosters, hens and chicks), and can effectively extract the group intelligence of chickens to Optimization. Starting from the goals of shortening time and limiting impact, trajectory planning in joint space is carried out through 5-degree B-spline curves to ensure that the entire trajectory is smooth while ensuring that speed, acceleration, and jerk are continuous. The chicken flock optimization algorithm is used to take the total time of movement and movement impact as the optimization goals, and comprehensively consider the process requirements and mechanical arm performance constraints. In the end, the high-speed movement of the robotic arm ensures smooth and stable torque changes during the movement of the robotic arm without sudden changes, and is suitable for trajectory planning from point to point. The CSO chicken flock algorithm brings advantages: 1. The movement trajectory is fast and stable; 2. The rhythm is improved.

Autonomous planning of motion trajectories

The Stial Rapid trajectory intelligent planning system integrates 3D digital model import (the unmanned navigation module can build its own model through the vision system), trajectory generation, simulation, trajectory adjustment, and code generation, solving robot programming problems in one stop; Stial Rapid trajectory
intelligence The planning system easily customizes the construction scene, and has powerful trajectory editing functions, such as: mirroring, array, trajectory reciprocation, unlimited addition of POS points in the trajectory, rewind editing at any step, etc.; the trajectory can be adjusted multiple times in the polishing application ;
The grinding process requires fast trajectory generation and multiple modifications based on the actual polishing effect.
Stial Rapid uses unique parametric trajectory design patent technology to parameterize every step you modify on the trajectory. When the modification results When you are not satisfied, you can go back to any step and modify the parameters, and subsequent steps will be automatically calculated, greatly improving the trajectory modification and iteration effects.
The Stial Rapid trajectory intelligent planning system also has an unmanned navigation function module. There is no need to import a theoretical 3D digital model. It can independently build a 3D model through the vision system and apply it to autonomous mobile polishing robots in large scenes.
Autonomous-planning-of-motion-trajectories

AI autonomous learning training evolution

The original Mogu Cloud system, based on deep learning, autonomous machine learning, reinforcement learning, domain data augmentation and domain generalization technology, perfectly solves the repetitive design and programming of workpieces in different fields, establishes a mapping relationship between polishing quality and process parameter settings, and
completely Unsupervised learning, the machine already has a cognitive structure similar to the human brain. In the subsequent upgrade and expansion, there is no need for humans to participate in formulating learning methods and defining rules. The machine can independently establish cognition and learning in the vast big data. Ability to intelligently formulate process routes for a wide range of working conditions based on sufficient data; achieve independent polishing.
And AI continuously improves its own behavior and decision-making through trial and error, self-calculates, self-corrects, self-plans paths, and shortens process debugging time.

AI-autonomous-learning-training-evolution

autonomous navigation

Autonomous navigation. This technology uses unmanned driving technology to allow the robot to have “sufficient” intelligent movement capabilities. An intelligent mobile collaborative polishing robot composed of an autonomous navigation mobile chassis + a flexible polishing robotic arm can realize the processing of large-sized parts (such as wind turbine blades). ) automatic polishing.
autonomous-navigation