News

Introduction of temperature curve machine learning for reflow ovens and AI for Sinhan's production process and management

Source:NexAIoT 

Introduction of temperature curve machine learning for reflow ovens and AI for Sinhan's production process and management

 

Background

NEXCOM is a world-renowned industrial computer manufacturer, covering industrial computers, network computers, digital surveillance, in-vehicle computers, digital multimedia and other related industries. In response to the trend of Industry 4.0, SHINHAN introduced its corporate digital transformation plan in 2017, planning to integrate its corporate headquarters, Sanmin manufacturing plant, and the new Hua Ya plant with OT/IT information to build an IoT architecture and corporate war room for Industry 4.0. In 2018, Hua-Ya's second factory was established in Taiwan, with a factory area of 8,000 square meters and two SMT lines, DIP lines and system assembly lines, providing highly automated OEM production of network communications and network security equipment.



 

Challenge

The invention and refinement of SMT (Surface Mount Technology) has contributed greatly to the booming of the electronics industry. Reflow is one of the most important technologies in surface mount technology, and the temperature rise curve of PCBs in the reflow oven will greatly affect the yield, while the current on-site equipment information is obtained by manual file checking and Key-In, which not only has higher labor cost, but also consumes a lot of time to adjust the parameters, and will have staff management problems, and manual transcription is also prone to distortion, resulting in subsequent production This makes it difficult to realize the application of intelligent manufacturing, such as management, production history, data analysis and production insight. 

 

Solution

- Edge Server:NISE3900R
- Software : PROFET AI Machine Learning Automation Solutions

With NISE3800 as the Edge Server and Profet AI machine learning automation solution, we can first input the product specifications (sheet size/thickness, production line, BOM, etc.) into the system and import the temperature rise curve of the reflow oven and other related characteristics and historical big data to build PCBA & Reflow production model. Afterwards, when a new product is launched, the new product specifications can be input and the AI machine learning will automatically get the best recommendation for Reflow parameters.

 

Results and Benefits

  1. PROFET AI simple and intuitive automatic machine learning platform not only a day to get started, a week on the ground, but also the previous human set Reflow temperature curve, converted by the past successful temperature curve data to do learning, reduce the staff Reflow tuning parameters time, but also significantly improve the Reflow yield.
  2. PCBA & Reflow production model retraining, customer Reflow production knowledge base management, production process and management AI, eliminating the problem of transferring experience from teachers, and using information systems to accumulate and preserve production processes.
  3. PCBA & Reflow production model retraining, customer Reflow production knowledge base management, production process and management AI, eliminating the problem of teacher transfer experience, and information system to accumulate and preserve the production process.

 

Site Language