On October 12, the 26th China International Industrial Exhibition opened at the National Exhibition Centre (Shanghai) in Shanghai. The theme of the exhibition was "Intelligent Economy, Intelligent Infinity", with an area of 300,000 m2, 10 major exhibition areas covering the entire industrial chain.
This is the eighth consecutive year that McCamand has made its debut at the Expo, which is located in the central exhibition area of robots (stand: 7.1H-A225) and has an area of 300 m2, focusing on 20 smart applications. Applications cover a wide range of robot shapes, such as industrial mechanical arms, humanoid robots, through the typical industrial smart scenery such as intelligent crawling, sorting, assembly, graphite, quality inspection, involving automotive, 3C / semiconductor, new energy, machining, logistics, and more than a dozen other industries.
The on-site exhibition applications are based on McMand's self-developed general eye-to-eye brainwashing system. By upgrading to full-chain eye-to-eye, brain-to-hand technology, robots are able to respond more steadily to complex tasks such as transparent objects, bulk SKUs, high-precision assembly, dynamic operations, and demonstrate cross-industry, cross-scenario, cross-task universality and generalization capabilities.
Based on the Mech-GPT multi-mode large model with a body brain at its core, the robot can understand natural language instructions and feed electronic labels, face hundreds of objects, such as food, everyday items, tools, toys, tools, and more, in real-time identifying object categories and adjusting sorting strategies; hand-to-hand operations with both arms and Mech-Hand to achieve precise grabbing and sorting of objects.
The unit demonstrates the ability of robots to generalize to unknown objects and open categories. By incorporating new upgraded eye-brain technology, the robot's grip speed and adaptability are further enhanced, allowing it to more flexibly cope with changes in position of objects, improve grip stability and rhythm.
By using the Mech-GPT multimodal large model, robots are not only able to understand multimodal information such as natural language instructions, images, but also to identify arbitrary objects, understand object relations, make logical and common sense reasoning, and perform complex operations such as classification, sorting, crawling, etc. according to task objectives. Natural language directly translates into physical movements, which greatly reduces the barrier to robot operation.
This is a multi-machine collaboration application in the field of human-made machines. Two human-made robots equipped with bracelets can autonomously collaborate to complete full process tasks, such as charging, assembly, charging, and transportation. The robot can identify multiple tools, understand global tasks, and complete route planning and movement execution, achieving efficient collaboration between multiple robots through the intelligent perception and AI planning capabilities of the human brain, combined with Mech-Hand precision operations.New generation of visual perception technology: ultra-high-speed dynamic image tracking
Dynamic point cloud graph (color of point cloud by height rendering)McCandless developed ultra-high-speed dynamic AI imaging technology that can accurately image objects moving rapidly from 100 to 500 mm/s, with a dynamic accuracy of up to 0.1 mm at 0.7 m. This high-precision dynamic visual perception capability can greatly improve the robot's ability to recognize and grasp high-speed moving animals, achieve edge tracking, edge positioning, edge operation, and significantly improve the production pace.
02
Flexible and precise assembly
Empowering high-precision smart manufacturing
High-precision assembly of FPC soft line
The FPC soft alignment is lightweight, soft and deformable, and the connector space is small, and the insertion accuracy requirement is very high. The unit uses the all-new robot 3D eye Mech-Eye UHP-040-5M, XY resolution up to 0.017mm, Z to repeat accuracy up to ±1μm, can clearly display micro-assembly characteristics, combined with AI recognition technology, precise positioning of the soft alignment endpoint and connector. The robot brain actually calculates the insertion position and angle, combined with the control of the self-contained motion to complete the soft alignment, and through the folding of the connector, achieves the complete process of plugging from the workpiece to the finished product, successfully working autonomously, with an insertion rate of more than 99.7%.Flexible grip and accurate assembly of power lines
Power lines are soft and deformable, while the need for stability and accuracy of the plug is extremely high, otherwise the plug cannot be inserted, the plug is not in place, or the plug force is too high to cause component damage. Based on the robot brain's task understanding and AI planning ability, the robot can automatically complete the automatic closure of the plug from the power line to the plug plug. At the same time, combined with high-precision vision sensing, adaptive motion control and precise operating capabilities, the robot can adjust the position, posture and power of the plug in real time, achieving the stability of the grip of flexible objects and the precision of the sub-millimeter level of the plug.High-precision welding and assembly of multi-specific metal parts
Based on the robotic brain's task-understanding and autonomous planning capabilities, the robot is able to recognize multi-specific metallurgical workpieces, accomplish crawling and assembly tasks autonomously, and combine 3D+2D visual synchronous perception with sophisticated operational capabilities to achieve sub-millimeter-level assembly accuracy. Currently, the application has been widely deployed in multiple process links in industries such as automotive parts, 3C electronics, and supporting flexible advanced manufacturing.03
Smart graphing and tunneling detection
Helping graphite quality control and process optimization
Eyeglass leg five-axis pinhole and tube detection
The unit is an all-new application introduced by McCandless to the spatial curved point curve scene. For the lens legs such as the curved shape of the curve, the curved curve should be attached to the curved curve scene, the General Eye + bracelet can accurately collect the three-dimensional contours of the curved workpiece, automatically generate the trajectory of the curve, and combine the five-axis synchronous control to achieve the multi-angle continuity of the curved workpiece. After the curve is completed, it automatically detects the quality of the graphite, achieving the integration of the make-up and curve detection.
Production interface mapPack box graphite detection & screening guides
Unit site and production interface map (slide to view)
Based on high-precision three-dimensional perception and intelligent analytical capabilities of the robotic brain, the robot can quickly obtain three-dimensional information about large pack boxes and surfaces, accurately measure parameters such as width, height, volume, and intelligent identification of surfaces such as overwidth, narrowness, and gaps. At the same time, the system can automatically generate surfacing paths, guide robots carrying a simulated surfacing gun along the trajectory, and autonomously complete surfacing path checks and air density checks.
Quality testing of car engravings online
Unit site and production interface map (slide to view)
The General Eye+Cerebral Palsy Test System can detect all types of defects in real time in the process of rifle grafting, such as fracture, bias, and leakage, detecting speeds of up to 1,500 mm/s, meeting the high-speed detection requirements; supports 360° full-angle detection, complex structural areas such as precise detection of angles, curves, holes, and improves the accuracy and stability of the detection of deformed workpieces; can produce detection results in real time, and supports historical data tracing and defect trend analysis, providing reliable data support for grafting process optimization.
04
High-precision 3D measurement and AI defect detection
Driving high-quality smart manufacturing
High measurement of electrosensitivity & detection of external defects
Unit site and production interface map (slide to view)
The Z-axis resolution of the Mech-Eye LNX-7030 is 1.9 μm and can be accurately measured at a height of only 2 mm; combined with dynamic ROI, multi-ROI, and other functions, the scanning speed is significantly improved, and the high-throughput can be measured at up to 0.5 s/s; while intelligent division imaging technology can accommodate transparent glass, three-dimensional electrode and reflective electrode materials, greatly improving the accuracy of detection and stability of complex scenarios.
Type-C connector visual defects detection
Unit site and production interface map (slide to view)
Through multi-site, multi-angle synchronous detection, a single site can complete all six aspects of the Type-C connector (including up and down planes, left and right curves, front and rear ports) with a look-all inspection task. Combined with Mech-DLK multi-model level and pixel-level fault division, it accurately detects dozens of types of defects such as pressure points, scratching wounds, kidney stains, deformations, electrical malformations, achieves fine, defective identification quantization and visualization presentation, and meets the requirements for high-precision component diversification quality inspection.
Laptop Heat Screw Height Measurement
Unit site and production interface map (slide to view)
It can effectively respond to interference such as dark surfaces and local shading, accurately extract the characteristics of the screw head, calculate the height of its relative reference surface, detect float height, locally unattached locks, and other installation problems. Through intelligent planning, scanning paths and multiple scanning pairing, it can cover scattered layout measurements, achieve high detection of multiple screw racks, and improve installation quality consistency.
Small multi-tasking motor flying quality inspection
Unit site and production interface map (slide to view)
The robot is equipped with 2D intelligent eyes, which take multi-point, multi-angle shots of small motors while moving. Combined with the analysis and detection capabilities of the AI brain, a single operation can complete tasks such as screws with no detection, geometric measurements, tag recognition.
Cell phone camera flash spot fault detection
Unit site and production interface map (slide to view)
For microfocuses of only 0.7 mm in size, the system combines images with multiple viewpoints in full-scale high-resolution images to obtain distinct focal point characteristics of different heights. Combined with pixel-level division and classification capabilities of AI deep learning, it accurately detects defects such as leaks, less ink, overlay, vacuum foam and foreign objects, and improves detection efficiency by taking high-speed shots with models in parallel reasoning to meet the high-speed detection needs of precision electronic manufacturing.
05
High-difficulty scene stable operation
Addressing industry challenges such as transparency/dynamics/high precision
Transparent objects widespread capture
Identification and crawling of transparent objects has always been a major challenge in the field of robotics. In this unit, based on self-developed transparent object AI imaging technology, and an AI brain built with mass-real data training, robots can effectively address high-permeability 3D imaging problems, generalized identification of common transparent objects. Combined with intelligent path planning and precise motion control technology, robots can independently plan crawling strategies based on object posture, and secure crawling of objects under complex processes such as random placement and disordered stacking.Transmitting dynamic to capture charges
Grabbing a moving animal body is a very challenging task for robots. Based on a smart eye-catching bracelet, the robot can effectively cope with core problems such as the continuous movement of the workpiece during the transmission belt supply process, position and gesture changes in real time, the movement of the workpiece, to achieve the precise positioning of the moving workpiece, tracking, grabbing and filling. At the same time, the system can help the robot adapt to the upper and lower positions of the multi-material transmission belt and the transmission nodes, helping the production line to achieve efficient, continuous automated material conversion.Black pentagonal screw high-speed disorderly charging
Small parts speed and disorderly charging are typical problems in manufacturing. In this unit, intelligent perception, planning and operation capabilities based on the eye's brain, robots can quickly identify and grab scattered stacks, small parts with black light-absorbing needles, and snaps up to 2.4s/piece, with a clearance rate of up to 99%.High-reflection turbines on high-precision machinery beds
Based on high-precision visual sensing technology, robots are able to accurately position and grab the workpiece, meeting the sub-millimeter level of the workpiece's precision requirements. At the same time, they can effectively cope with problems such as high reflection, postural randomness, oil pollution interference, and high-precision machine automation of waste.High-precision car parts mounted on racks
In this unit, through 3D+2D visual synchronization, the integrated positioning accuracy can reach ±0.5mm. At the same time, AutoFit AI technology, such as multipoint fit, can adapt to changes in the shape of the suspension, automatically compensate for position biases, achieve precision parallel suspension, and a success rate of over 99.9% for suspension.
In this exhibition, McCamand, together with several heavyweight partners, presented Mech-GPT multi-mode large-model human interaction, flexible precision assembly, silicon gold processing, and other applications for smart eye brains.Guides to the exhibition
Exhibition time: National Exhibition Centre (Shanghai) Location: 7.1H-A225.
Welcome to the McCandless standAt this trade fair, McMand will continue to enhance the universality, reliability and ease of use of its products, work with industry partners, and create replicable and scalable application prices in more industries and scenarios.
From October 12 to 16, we invite you to attend the 7.1H-A225 McCandless exhibition at the Shanghai National Exhibition Centre, where you will experience the cutting-edge technology and rich applications of the general eye bracelet in close proximity!
* Translated from McCandless
RoboInd editorial channelResearch, robotics companies and market intelligence for cross-border sourcing.Explore the channel ↗
Share this
End of editorialExplore more from Global Robotics News ↗