DogwoodAI has utilized appropriate machine learning technologies based on client requirements and the characteristics of the factory, developing AI solutions through various approaches as outlined below that enable an autonomous manufacturing plant.

Tailor-made ML Technology

Tailor-made ML Technology adequately reflects the demands of clients in the industrial field. It predicts real-time multiple products based on data from various sensors (RTDB) linked to chemical plants.

The technology is developed by analyzing and cleaning big data generated in the field and then applying it to machine learning.

A Korean oil company has already conducted field verification tests of this technology for 6 months to ensure its reliability.

UI Platform for Tailor-made ML for HCR

Real data-based ML Technology

Real data-based ML Technology can analyze real-time data emerging from manufacturing plants, providing data insight, and enabling real-time predictions of main process variables such as product yield and properties.

Compared to conventional platform AI, this technology boasts superior predictive capabilities, offering advantages that can be utilized to enhance productivity and reinforce operational safety.

Real data-based ML Solution for Power Plant

Model-based ML Technology

Manufacturing processes involve the interaction of various units and devices. When dealing with raw materials, reactions, and products, first-principle mathematical equations can be employed to create models that predict results in simplified scenarios. These models can generate generic data and conduct sensitivity analysis for machine learning applications after validating the mathematical models.

Our technology showcases a machine learning model based on a dynamic model of an integrated process for fuel cell- grade hydrogen production. This model facilitates the recovery of hydrogen and the capture of carbon dioxide from tail gas produced by steam methane reforming. We have validated the data generated by this dynamic model using pre-existing data, thereby demonstrating the efficacy of model-based machine learning for developing AI and machine learning solutions in processes with established mathematical models.

This breakthrough represents a significant advancement in the field of machine learning for integrated processes, as it enables the creation of AI and machine learning solutions for scenarios where mathematical models can be constructed.

Model-based ML Solution for the Integrated process

Physics-based ML Technology

The use of AI software in manufacturing processes has become increasingly important. However, obtaining the necessary data to develop these solutions can be challenging. This is because such data is not always collected in the industrial field. To address this issue, DogWoodAI have developed different ML technologies that combine scientific theory and machine learning.

Physics-based ML Technology is one such technology that merges physicochemical theory with big data machine learning. This approach enables the reproduction of consistent performance beyond the range of actual use and the prediction of future data.

Prediction of solubility in electrochemical solutions through Physics-based ML Solution

AI Software

Sub menu

Tailor-made ML Technology

Real data-based ML Technology

Model-based ML Technology

Physics-based ML Technology

AI Software

Sub menu

Tailor-made ML Technology

Real data-based ML Technology

Model-based ML Technology

Physics-based ML Technology

개인정보보호정책

이용약관