Product description:

Core Technology: By incorporating AI deep learning algorithms, artificial neural networks and multi-layer convolutional kernels into the coating inspection system, the application of AI algorithms enables the Raise-A8 to achieve intelligent, minimalist programming. Furthermore, AI deep learning facilitates intelligent decision-making, thereby resolving the two main shortcomings of traditional algorithms: lengthy programming times and high false alarm rates.

Image:

Sepcification:

Category	Items	Details
The software-related	operating system	Windows 10 64-bit system
	The communication method	SMEMA interface
	Machine vision algorithms	feature extraction, morphological calculation and analysis, classification and discrimination, and color conversion mode "Version matching OCR, etc.
	Deep learning algorithm	rtificial neural network + multi-layer convolutional kernel + reinforcement learning algorithm (self-developed)
	The inspection speed	280-320ms/FOV
	The learning quantity	greater than 100 FOV images
Electrical requirements	Power supply and power	AC220V, rated power 3000W
	The gas source	0.4-0.6Mpa
	Working environment	temperature：10~45℃,humidity：30~85%RH
D coating thickness measurement	principle	Coaxial confocal laser sensor + Spectrometer (self-developed)
	The measurement range	30~1000um
	Accuracy:	3um
	The measurement time	approximately 1.5 seconds per point (including mechanical movement time)
The function configuration	multi-glue and low-glue detection	determine whether the coated area is normal
	Bubble detection	detect bubbles with a diameter of 100um, display their positions and count the quantity
	Splash detection	statistics cover the area of scattered coating points
	Wrinkle detection	etects wrinkle defects caused by uneven curing energy
	Reverse wetting detection	used to detect reverse wetting defects caused by excessively low dyne values on the board surface
	Aare detection	The combination of thickness and area detection can determine the amount of glue
	The glue width detection	supports any shape such as polygons and circles Array images are suitable for imposition processes
	Array images	suitable for imposition processes
	Scan the code	identify and confirm the identity of the PCBA
	Intelligent path planning	automatically identifies key detection areas through deep learning algorithms
	SPC statistical	data statistics and data traceability functions
	MES communication	customized according to requirements
PCBA specification	dimensions	50x50mm~550*650mm
	Component height	Top surface：60-90mm,bottom surface：120mm
	The edge of the process	3mm
	The weight of the PCB board	15 kilograms (including the carrier)
Hardware	Overall size	1130mm×1510mm×1700mm
	Overall weight	800kg
	XY motion mechanism	grinding lead screw + servo motor
	The track width	Electric
	The track height	900mm±50mm
	The movement speeds of X and Y	0-1000mm/S
	Running direction	left in and right out (optional right in and left out)
Optical Specifications	Camera	25MP color industrial area array camera
	The lens	a telecentric lens with a depth of field up to 200mm
	Light source	UV+W light source (Customized UV wavelength)
	FOV	42*42mm(Optional 32*32mm)
	Pixel accuracy	8.3 μ m(6.2 μ m optional)
Industrial control host	CPU	the industrial control host is the 13th generation Intel processor with a frequency of 5.4GHz
	Memory	DDR55600MHZ128G
	Hard Disk Drive	2T
	VGA Card	Nvidia Discrete GPU Accelerator Card (4060Ti8G)
	Network card	Gigabit network card