Introduction～Characteristic Value Prediction～

In the manufacturing industry, there are cases of trial and error of various setting values when developing products. For example:

• Optimize the mixing ratio of multiple raw materials during material development to achieve the desired properties (Hardness, etc.)
• Adjust manufacturing equipment and design settings during product development to achieve desired characteristics
• Tune several parameters to get good software performance (Power consumption, etc.)

If the measurement of the characteristic value requires cost and time, the above trial and error will take considerable time.

Predictive analytics enables you to predict a property value for a setting (Example: Mixing Rate, Settings, Parameter Settings) from historical trial and error data. By using this function, you can determine the setting that results in a higher characteristic value. Therefore, trial and error may be promoted more efficiently than trial and error based on experience and intuition.