As technological advancement of machine learning continues, the key to a successful digital transformation lies within the ability to build a data driven prediction model. AMR allows users to experience a self-evolving prediction model generation process of building the most appropriate model through sequential sampling even without the expertise on sampling, metamodeling, or machine learning.


Outlier Filter

Effectively removes outliers (minority of erroneous samples) to prevent them from damaging model accuracy.

ML Model Generation AI

Uses PIDOTECH’s exclusive AI-based prediction (Bruce) to generate an accurate prediction model based on a robust set of data.

Rule-based Sequential Sampling Manager

A state-of-the-art sampling method couple with PIDOTECH’s know-how generates an elaborate prediction model using minimum amount of sample points.

ML Model Export Manager

Exports prediction models in various formats (Excel spreadsheet, PIAnO model, executable, etc.) for a wider accessibility.


Wider accessibility

Lowers the barriers to entry for prediction model generation as it no longer requires background knowledge
(design of experiments, metamodeling, prediction error analysis, etc.).

Man-hour reduction

The entire process from data generation to prediction model building is done automatically.

Establish big data generation process

As AMR is an automated data generation process, it can be utilized as a system for data accumulation and ultimately be used for data-driven design.