Project

Tailor-made AI solution for a hydrocracking plant in a refinery company

Challenge

Refinery companies require highly accurate, real-time predictions of product yield and quality in complex hydrocracking processes to optimize operations and reduce variability.

Approach

DogWoodAI developed tailored-made AI model designed specifically for hydrocracking plants. The model was built to predict product properties within 1% error and was implemented in close collaboration with the client to ensure continuous field monitoring and improvement.

Value Delivered

The solution was validated over a six-month field test in a Korean refinery. It demonstrated highly accurate, real-time prediction performance, leading to improved process control and decision-making capabilities in a mission-critical environment.

Tailor-made ML solution for a hydrocracking plant

Hybrid Simulator based on machine learning and mathematical model for the fine chemical production

Hybrid Simulator for a fine chemical plant

Challenge

Fine chemical plants face limitations with conventional mathematical models in handling process complexity and often lack sufficient data for training standalone AI models.

Approach

DogWoodAI developed a Hybrid AI Simulator combining physics-informed mathematical models with data-based AI techniques. This hybrid approach was co-developed with a global platform partner to complement the strengths of each methodology.

Value Delivered

The hybrid simulator, achieving prediction deviations of less than 0.5%, was deployed in a domestic fine chemical company. It provides accurate real-time predictions of product quality and yield.

Hybrid Simulator for a fine chemical plant: Interview with AVEVA

Challenge

Traditional commercial simulators are often limited in capturing real-time plant behavior and delivering predictive performance in complex chemical processes.

Approach

In collaboration with AVEVA, DogWoodAI integrated AVEVA’s first-principle simulation capabilities with AI model, forming a hybrid simulation platform tailored for chemical plants.

Value Delivered

The project successfully demonstrated the potential of Hybrid Simulation to outperform conventional commercial simulators. The collaboration opened new possibilities for scalable adoption across diverse manufacturing sectors, showcasing DogWoodAI’s innovation in applied industrial AI.

Computational Fluid Dynamic models for three-phase reactor in a fine chemical company

Challenge

Optimizing mixing and reaction performance in three-phase reactors is complex due to the interactions among gas, liquid, and solid phases, especially during solid particle formation.

Approach

DogWoodAI applied advanced CFD modeling to simulate flow patterns and phase interactions within the reactor. The study focused on determining the best impeller design and operating conditions to maximize mixing efficiency.

Value Delivered

The project identified optimal reactor configurations and operating parameters, laying the foundation for the development of a CFD-based AI model for future predictive simulations and performance optimization.

CFD results for the mixing performance of the reactor (Gas-Liquid-Solid phases)

Consulting Project: Xenon recovery process

Challenge

Designing a scalable and efficient process for Xenon recovery from Helium mixtures requires accurate process modeling and simulation based on limited experimental data

Approach

DogWoodAI developed a model-based framework that included constructing a physicochemical database, creating a validated breakthrough simulator, and performing process simulations to assess feasibility.

Value Delivered

The project confirmed the technical feasibility of the Xenon recovery process and provided key design guidelines for scaling. It also demonstrated DogWoodAI’s strength in combining theoretical modeling with real-world validation for future AI-driven development.

Xe recovery process results

Advanced AI Solutions for Fine Chemical Plants: DogWood Autonomous Manufacturing

Challenge

Fine chemical plants require intelligent systems capable of handling dynamic changes, such as catalyst aging, while ensuring product quality, operational safety, and maintenance efficiency

Approach

DogWoodAI launched the DogWood Autonomous Manufacturing, an integrated AI framework featuring real-time data analysis, anomaly detection, optimization guidance, and safety insights. The system processes data for each tag in real time to predict product outcomes.

Value Delivered

The AI system has been operational in a fine chemical plant, offering real-time predictive insights, anomaly alerts, and data-based maintenance reports. It significantly enhances production stability, safety, and operational efficiency.

DogWood Autonomous Manufacturing

Development of an AI-driven simulator for semiconductor material batch process (ongoing)

Challenge

Due to the nature of the semiconductor material batch process, multiple operation and analysis cycles are required to achieve target product quality. Reducing these repeated operations through accurate quality prediction is a key challenge.

Approach

DogWoodAI is developing an AI-driven simulator that supports intelligent decision-making by analyzing process data using domain knowledge and training AI models to predict accurate product quality for each batch operation.

Value Delivered

Through accurate predictive simulations, the solution significantly reduces unnecessary process iterations, thereby maximizing production efficiency and yield. In turn, this leads to cost savings and enhanced process stability, strengthening the manufacturing operation’s overall competitiveness.

개인정보보호정책

×
다양한 팝업 용도로 사용하세요! 여러개 등록 가능합니다!

바로가기 버튼

이용약관

×
다양한 팝업 용도로 사용하세요! 여러개 등록 가능합니다!

바로가기 버튼