In this project, we developed and implemented a Surrogate Reservoir Model (SRM) to optimize field development strategies in a complex oil reservoir. Traditional full-physics simulations, while accurate, are often too time-consuming for rapid decision-making and scenario testing. To address this, we built an AI-powered SRM that replicates the dynamic behavior of the reservoir with high accuracy but at a fraction of the computational cost. By training the model on a set of carefully selected simulation scenarios, we were able to create a predictive system capable of forecasting reservoir fluid dynamics under different well configurations and operational strategies.

The results of this study demonstrated the practical value of SRM in industrial workflows. It enabled faster analysis of multiple development plans, supported proactive decision-making, and significantly reduced simulation runtimes. The approach is particularly beneficial for real-time optimization, production forecasting, and uncertainty analysis in mature fields or data-rich environments. This successful case study highlights how combining domain expertise with AI technologies can improve reservoir management efficiency and support smarter, data-driven field development in the oil and gas industry.

In collaboration with China University of Petroleum (East China), Shandong University of Science and Technology, and China University of Mining and Technology, this project aimed to revolutionize microporosity detection in tight sandstone reservoirs using deep learning techniques. Traditional methods of analyzing scanning electron microscopy (SEM) images often fail to capture the full scope of micropores due to resolution limitations and manual bias. To overcome these challenges, we developed a custom deep learning model (SdstMcrp) based on a U-Net architecture with a ResNeXt50 backbone, trained on thousands of annotated SEM images from the Beibuwan Basin.

The model achieved up to 60% improvement in micropore segmentation coverage and higher accuracy in permeability prediction using fractal analysis. This technology provides a practical, automated alternative to time-consuming manual interpretations, enabling faster digital rock analysis, improved reservoir quality assessment, and enhanced input for reservoir simulation and production planning. The project demonstrates how AI-driven workflows can be effectively deployed in core analysis labs and field development planning for unconventional oil and gas resources.

To address growing complexity in EPC operations within the telecom sector, our team collaborated with Parto Abi Group to develop a fully customized project and cost control platform. The system was designed using Microsoft Access and VBA and was deeply integrated with the company's existing Arian System financial software. This integration enabled the automatic synchronization of project and financial data, allowing detailed tracking of budgets, execution methods, cash flow, and real-time liquidity needs. The platform also featured built-in analytics for cost deviation and performance evaluation, helping project managers stay aligned with financial planning. 

The implementation of this system led to significant operational improvements. Manual data entry and Excel-based tracking were fully replaced by a centralized tool, reducing errors and improving consistency across departments. Financial managers gained timely insight into project-related expenditures, while technical teams benefited from enhanced visibility over cost flows and timeline tracking. Ultimately, this digital solution empowered Parto Abi Group to streamline project execution, enhance cross-departmental collaboration, and make faster, data-driven decisions in a highly competitive sector.

In 2023, we partnered with Pat Roshan Nikta, a major refractory materials manufacturer, to implement an advanced Business Intelligence (BI) solution tailored to their enterprise needs. The system was developed using Microsoft Power BI and fully connected to the company's PostgreSQL-based ERP platform (Odoo). Our team designed a suite of real-time dashboards covering finance, HR, and project management, allowing stakeholders to monitor KPIs, cost trends, and performance metrics across the entire organization. All dashboards were developed in close alignment with the company’s reporting structure and executive decision-making needs. 

The result was a transformative shift in how the company engaged with its data. Instead of static, delayed reports, managers now had on-demand access to live insights, making it easier to detect anomalies, optimize spending, and track project outcomes. The BI system fostered a culture of transparency and accountability by bringing data into every key discussion. By the end of the project, Pat Roshan Nikta had not only modernized its reporting infrastructure but also empowered its leadership with a decision-support system that scaled with business growth.