Products and Solutions

SMore Factory Intelligent Manufacturing AI Solution

SMore Factory provides full digitalization solutions for the manufacturing process, including production, quality control, equipment maintenance, etc.
It also provides a full-stack and one-stop AI empowerment and management ability, creating a standardized AI delivery system for intelligent manufacturing.
Product Introduction

Product Introduction

SMore ViMo (Vision Inspection and More) is an end-to-end vision AI system for intelligent manufacturing, designed and developed by two major engines: model training and model inference.

The AI model training system autonomously conducts algorithm learning for complex scenarios throughout the production line, including material tracking, defect locating, product quantity counting, multiclass appearance defect detection, etc. Once the algorithms from the model training engine are imported, the model inference engine can be deployed to the production line immediately, achieving real-time AI inspection.

4 Core Algorithm Functions

A new generation of vision AI technology, backed by 20+ years of practical experience
SmartMore gathered insights from over 20 years of research on AI technology to develop SMore ViMo, which provides four core vision AI algorithms functions for manufacturing, targeting significant difficulties such as the large variety of the products and the high frequency of the upgrade.

OCR
OCR

The end-to-end solution supports single-character and multi-character labeling and recognition, breaking traditional methods' limitation, solving complex issues such as curved surface character recognition, low-contrast character recognition, large character recognition, etc.

Inspection
Inspection

The technology of locating and classifying the targets in the detected materials can be used for various purposes, including multi-target detection, small target detection, counting, etc. It applies to scenarios such as counting drug pills and 3C device detection.

Classification
Classification

The algorithm can classify the inspected materials and run inspection tests such as the OK/NG test, object color detection, food material classification, and 3C product defections fine-classification, etc.

Segmentation
Segmentation

The algorithm can conduct detection and edge recognition on the detected material at a pixel level, such as identifying small cracking areas in a silicon wafer, and damaged areas in bearing's, etc.

Product Features

  • Consumer-level Using Experience
  • Clear Test Results
  • Visualized Training Process
  • Flexible Project Delivery
Consumer-level Using Experience

Consumer-level Using Experience

The product provides data labeling, management, and other functions with a user-friendly UI. Annotation work can be completed smoothly through guided labeling and quick closure. The platform supports import and export of the labeled data and images, suitable for further sharing and management.

Clear Test Results

Clear Test Results

Once the model training is completed, SMore ViMo will move on to the model testing. The testing provides model information, testing standards and visualized images, making it easier for users to quickly identify the functionality of the model. The testing report can be exported with one-click for users to further analyze and summarize the data.

Visualized Training Process

Visualized Training Process

Features such as automatic parameter tuning and intelligent data distribution mean that users do not need to have any professional AI knowledge, as they only need to perform simple parameter input to start the easy (one-click) training. During the model training process, the system provides the tendency curve of the model's results in real-time, visualizing the functionality for users to observe.

Flexible Project Delivery

Flexible Project Delivery

The model can be easily deployed to online inspection platforms, and the inspection results are displayed in real-time. The platform can dock with multiple camera data interfaces. It also supports multi-view display. After a few simple instructions, the model can be applied to the production line.

Application Scenarios

1

The OCR recognition module identifies the license plate of the vehicles carrying incoming goods. It can scan and recognize boxes, as well as recognize the material's ID while counting its number.

2

The materials are classified by AI to remove those that cannot meet the production requirements. Therefore, it can control the production quality at the very beginning of the production line to reduce production losses.

3

During production and assembly, by introducing AI, the system can randomly grab objects, conduct goal guidance, and achieve automated assembly. It can also carry out 3D detection on production modules and reduce the risk of producing defects.

4

After assembly, AI can be used for defect detection to control product quality and reduce the number of unqualified products that may flow into the next procedure. The data collected form each step of the production line provides valuable insight into the capacity in real-time, giving early warnings to production managers to adjust production dynamically.

5

As the product is now ready for departure, OCR can scan the product batches to keep track of their logistics, connecting the delivery and the aftersales procedure. This process ensures the information is organized, digitalized, and documented for further analysis.

Applicable Industries

Consumer Electronics

Consumer Electronics

Automobiles

Automobiles

New Energy

New Energy

Pan-industry

Pan-industry