The SMT stage in electronics manufacturing is the most complex one.  High levels of automation are required as well as technical know-how. Success in the SMT process will insure:

  • Reliable products 
  • High yield 
  • Manufacturing efficiency


Failure to establish a good SMT process results in:

  • Products of poor quality with a high rate of customer returns
  • Low yield and low production capacity
  • Additional labor and component costs  

In general, companies focus on setting up a good SMT system by focusing mainly on the testing equipment (AOI, SPI, ICT, Flying Problem, visual inspection, etc)

Do you think focusing on feedback from test equipment is enough? According to our experience with hundreds of  SMT lines, 95% of companies still struggle with achieving an efficient SMT process. What are the main causes of poor SMT processes? There are three main reasons:

  1. Bad design for manufacturability. 
  2. Wrong set up of the SMT process (machines, oven and solder paste). 
  3. Bad components (out of SPEC, fake, defective and so on). 

Despite the importance of the above three causes, the most effective means of setting up a successful SMT process comes from a different perspective:

We can only significantly improve SMT if we can unify all data from each stage and correlate the causes and problems quickly and accurately, creating a complete end-to-end control system.

Could this be possible? In SMT, how can we unify and correlate the various data sources if they are all compiled and stored differently?

Using Artificial Intelligence (AI) technology, automate the SMT process by correlating and analyzing every type of SMT data.

Artificial intelligence can be a significant benefit to the manufacturing industry. SMT process control can be integrated into a holistic end-to-end process to:

  • Increase yield and capacity. 
  • Improve the reliability and quality of the assemblies. 
  • Predict failures 
  • Reduce waste loss
  • Make better decisions and enhance the quality of manufacturing processes.

AI technology has solved IT teams’ biggest challenge: synchronizing all data from all plants. Machine data, sensor data, manual data, or any other data gathered from multiple sources can be processed by artificial intelligence that runs multivariable algorithms for big data from different locations to optimize operations.

SMT yield and quality can be improved more quickly by using artificial intelligence to correlate the diagnostics of the process. Monitoring, diagnosing, and tracing data will prevent faulty solder joints, defects, and costly repairs. By automating and optimizing the entire SMT process with AI, a more intelligent inspection can be achieved.

smt quality
The use of artificial intelligence to improve SMT yield and quality

QualityLine AI Analytics- Specialized for SMT process

As an expert system for the electronics manufacturing industry, QualityLine has a module devoted to the analytics of SMT processes. Using its automated AI data integration module, the system can integrate pick & place, AOI, flying probe, SPI, ICT and repair/rework data. Using QualityLine, data from multiple sources can be integrated into a unified, digital twin database to optimize SMT manufacturing efficiency and product quality.

Digital twin
Ai analytics

QualityLine -SMT Analytics root cause analysis: 

  • Prediction of failures 
  • Automatic alerts.
  • Automatic reports 
  • Testing stations drill down and analysis
  • Retests drill down and analysis
  • Repairs drill down and analysis 
  • Preventive Maintenance of testing equipment. 
  • Traceability and test history of every serial number through its life cycle.

Learn more at our webinar next week March 30th on how to improve SMT yield and quality by correlating the diagnostics of SMT, AOI, Flying Probe, X-Ray, Functional Testing, repair data, etc. Register here