#AI CROSS #Deep Predictor #IKO International

AI-Driven Demand Prediction Enhances Efficiency

In a transformative leap for operational efficiency, AI CROSS, a company based in Minato, Tokyo, has successfully introduced its AI demand prediction and management service, Deep Predictor, at IKO International, a subsidiary of Japan Thomson operating in the U.S. This implementation has revolutionized IKO's inventory ordering operations, slashing the time spent on these tasks from 3.8 hours to just 1.4 hours per week, which amounts to an impressive 63% reduction or approximately 124.8 hours annually. This case study is available on the official AI CROSS website.

Manufacturing Sector Challenges

The manufacturing industry faces ongoing structural challenges, particularly concerning labor shortages, the individualization of tasks, and the increasing complexity of supply chains. As highlighted in the 2025 Manufacturing White Paper by Japan's Ministry of Economy, Trade and Industry, the workforce in manufacturing is expected to decrease from 10.55 million in 2023 to 10.46 million by 2024, exacerbating existing issues related to labor shortages. To maintain and enhance productivity with limited personnel, the report emphasizes the need for breaking free from individualized tasks and leveraging digital transformation (DX) and AI to standardize and visualize operations.

Moreover, manufacturers are increasingly pressed to adapt to low-volume production of a wide variety of items, prepare for supply chain disruptions, and streamline operational efficiency in overseas locations. Particularly in the supply of critical machinery parts like bearings, businesses must navigate the balance between preventing stockouts—which can halt production—and managing the financial burden of excess inventory.

Responding to these industry-wide challenges, AI CROSS offers a solution with its Deep Predictor service, facilitating both the standardization of personalized operations and the enhancement of multi-branch efficiencies simultaneously. This release showcases IKO International's efforts as a prime example of the service's implementation.

Background of the Implementation

IKO International supplies mechanical components, such as bearings and precision machinery, to OEMs (Original Equipment Manufacturers) and distributors in the U.S. market, operating multiple sales locations across New Jersey, California, and Illinois. Traditionally, inventory ordering at these sites was conducted manually through Excel spreadsheets, leading to several challenges:

• - Operators had to manually determine order quantities per part number, resulting in a collective workload of 3.8 hours per week across four personnel.
• - Variations in order quantities depending on individual expertise led to excessive stock due to overlapping orders between sites.
• - The necessity of balancing stockouts and excess inventory remained a pressing management issue, given that bearings are critical components used across various industries.

Additionally, the transition to a new core system made the existing manual methodology unfeasible, highlighting the urgent need for an updated inventory ordering system.

Impact of the Implementation

Quantitative Effects:

Qualitative Effects:

• - Resolved variations in order quantities and eliminated overlapping orders across sites.
• - Achieved standardization of ordering operations, reducing reliance on individual personnel expertise.
• - Established a robust foundation for inventory management compatible with the new core system.

Feedback from IKO International

According to Asukata, a representative from IKO International, “Deep Predictor covers everything from demand forecasting to recommended order quantities in one service, effectively aligning with our operational reality. Additionally, with large-scale investment decisions often being a hurdle, the cost-effectiveness of Deep Predictor was significantly a factor in our choice. Post-implementation, we have achieved standardization of our previously personalized inventory ordering tasks and drastically reduced working hours across all branches in the U.S.”.

About Deep Predictor

AI CROSS's Deep Predictor service embodies the principle of ‘Enabling anyone to easily make predictions.’ This service aims to assist on-site operators—regardless of their expertise—in utilizing existing data to make high-accuracy predictions and informed decisions. it’s not just a forecasting tool but generates outputs directly relevant to business operations, thus designed as **