Arshin Taghtamish, Contract and Equipment Procurement Manager at SSMIC, stated in an interview with the Public Relations correspondent:

Sangan-Khorasan Steel Mining Industries Company (SSMIC) has taken a major step in optimizing its supply chain processes by leveraging advanced technologies. Through a project titled “Intelligent Supplier Entry and Allocation Process Using Data Mining and Machine Learning Techniques”, the supplier entry and allocation processes were designed and implemented using advanced data mining algorithms and machine learning models.

Referring to the key objectives and outcomes of the project, he added: The main goal was to enhance accuracy, transparency, and efficiency in supplier selection and reduce reliance on traditional methods and subjective judgment. Among the notable achievements are: a significant reduction in evaluation errors, smart monitoring of active suppliers, development of a pilot software for full integration with current systems in future phases, and step-by-step documentation of the intelligent sourcing roadmap in the steel industry.

Taghtamish further explained: Based on this study, it will be possible to scale the intelligent sourcing system, integrate it with other parts of the supply chain, and utilize technologies such as the Internet of Things (IoT) and intelligent decision support systems. This initiative marks the first step in SSMIC’s supply chain digitization roadmap, which includes stages such as establishing digital foundations, advanced data analytics, systems integration, and the development of advanced artificial intelligence.

Referring to the long-term objectives of the project, the Procurement Manager noted: The long-term goals include cost reduction, increased transparency, and positioning SSMIC as a leader in intelligent sourcing within Iran’s steel industry. This initiative is considered a strategic step towards digital transformation and enhancing the company’s competitiveness both nationally and internationally.

Introducing the Intelligent Supplier Evaluation Software

Arshin Taghtamish, Contract and Equipment Procurement Manager at SSMIC, explained the features of the newly developed software:

The software is based on one of the key techniques in machine learning—Support Vector Regression (SVR)—and is designed to evaluate suppliers. It can automatically assess suppliers in both the entry and allocation phases using various input variables and compare results with traditional expert-weighted methods.

In its initial implementation, the software has significantly improved the efficiency and accuracy of the evaluation process. It enables the company to automatically assess suppliers using reliable data and helps eliminate subjectivity from decision-making.

He went on to highlight the key features of the software:

Inputs and Data Sources:
The system automatically ingests multiple variables, including supplier characteristics (such as financial capability, supply history, resources and support, standards, and localization status), as well as output data from previous periods (responsiveness scores, technical scores, bid-winning scores, quality control scores, and on-time delivery scores).

Predictive Model:
The software uses the Support Vector Regression (SVR) model with an RBF kernel to predict supplier evaluation scores. It processes various supplier features as input and generates predicted evaluation scores as output.

Outputs:
The system calculates five distinct scores for each supplier, as well as an overall supplier score representing the general performance evaluation.

Comparison with Previous Method:
Previously, supplier evaluations were based on expert-weighted scoring and subjective assessments. Such methods were prone to inconsistencies and bias. The new software enables more accurate, faster, and objective evaluations using mathematical models and machine learning techniques.