"Finding Optimal Components with AI"... LG Innotek Builds 'AI Component Proposal System'

Integration of 2 Million Internal and External Component Data Sets

Component Selection Streamlined with AI-Driven 'Reference Price' Feature

Quotation Time Cut by 70%, with Enhanced Proposal Quality

LG Innotek has established a system utilizing artificial intelligence (AI) to quickly identify the optimal components for new products, aiming to boost its competitiveness in securing orders.

On September 20, LG Innotek announced that it has built the 'AI Component Proposal System,' which analyzes internal and external component data using AI to allow users to compare specifications and prices for each component. After more than two years of development, the system was recently rolled out across all business divisions.

LG Innotek’s products, such as camera modules, semiconductor substrates, and automotive components, typically contain hundreds of individual parts. Previously, calculating quotations for bids and developing products required checking data scattered across multiple internal systems and investing significant time in discovering new components. By introducing AI into the component search process, LG Innotek aims to reduce the related time, select the most optimal components considering both price and performance, and thereby enhance the quality of customer proposals as well as its competitiveness in winning contracts.

For the development of this system, LG Innotek consolidated data on approximately 2 million components, including capacitors and inductors, sourced from both within and outside the company. Component names, specifications, and units—which vary by manufacturer—were standardized to build a database usable by AI, forming the foundation for the new system.

LG Innotek 'AI Component Recommendation System'. LG Innotek.

When a user enters the required component specifications, the AI scans the entire database to return similar parts. Not only does it search components previously purchased by the company, but it also covers external market data. This expands the scope of parts searches far beyond the prior approach, which depended on individual experience and purchase histories.

The key differentiator of the system is its AI-based 'reference price' estimation feature. The AI analyzes purchase histories and price fluctuation trends of similar components and proposes a current price point. The reliability of these reference prices surpasses 96%, enabling accurate price forecasts for both previously purchased components and new parts without prior purchase records.

This allows a shortlist of price-competitive candidate components to be selected in under two hours. Actual quotations can then be confirmed with shortlisted suppliers, making the selection of components far more efficient. Industrial components, unlike consumer goods with set retail prices, vary in price depending on factors such as purchase volume, contract terms, and timing. Previously, finding price-competitive parts required requesting quotes from multiple suppliers and conducting extensive market price research.

An LG Innotek representative explained, “Recently, AI-driven component recommendation systems have emerged one after another, but our implementation of a 'reference price' calculation feature for each part is the biggest point of difference from existing systems.”

With the introduction of the system, the time required to generate new product quotations has been reduced by more than 70% compared to before. LG Innotek expects to utilize this saved time to more thoroughly review customer requirements and improve proposal quality, thereby further enhancing its competitiveness in winning orders. The company will also be able to flexibly respond to supply fluctuations by rapidly securing alternative parts if a specific component is discontinued or an interruption occurs in supply.

In recognition of these innovative achievements, the system won the Customer Satisfaction Award at the '2026 LG Awards' held in April. At present, AI searches for components and calculates prices using previously collected and standardized internal and external parts data. In the future, LG Innotek plans to further develop the system by incorporating agentic AI, which will autonomously discover and analyze the latest component information and update the system accordingly.

Junseong Kim, Head of the Purchasing Center (Executive Director) at LG Innotek, stated, “The ‘AI Component Proposal System’ is a meaningful innovation that completely transforms existing practices, tailored for manufacturers dealing with diverse components. We will continue to provide value that exceeds customer expectations through AX (AI Transformation)-based working methods.”

LG Innotek is also accelerating AX by introducing AI throughout its operations. The adoption of 'AI Incoming Raw Material Inspection' in key production processes has reduced the time needed to analyze causes of material defects by up to 90%. The application of 'AI Vision Inspection,' where AI detects defective finished products, has improved yields for camera modules and semiconductor substrates.

The application of 'AI Process Recipe' has reduced the time needed to determine optimal processing conditions for camera modules from 72 hours to within 6 hours. According to the company, its 'Dream Factory'—an FC-BGA (Flip Chip Ball Grid Array) substrate production hub utilizing AI and robotics throughout the production process—is considered the industry’s leading smart factory.

Recently, LG Innotek also applied the 'EXAONE Tabular' industrial AI model developed by the LG AI Research Institute to its production sites. This has cut the time required for AI to learn changed process conditions by approximately 85%, further strengthening group-wide AX collaboration.

Additionally, LG Innotek was the first among LG affiliates to introduce an in-house AX certification test for employees. More than half of office staff have obtained certification, accelerating the internalization of AX capabilities across the company.

This content was produced with the assistance of AI translation services.