#ZKH Group #China #Suzhou #Industrial AI #LingLong Intelligence

ZKH and Intel's Milestone in Industrial AI

In a groundbreaking announcement, ZKH Group Limited, a prominent name in the maintenance, repair, and operations (MRO) sector, has collaborated with Intel to introduce the LingLong Insight 1.0. This innovative multimodal AI model aims to revolutionize industrial operations with improved verification capabilities, making it easier for workers to manage and confirm their supplies on-site.

The Challenge at Industrial Worksites

The industrial landscape is often characterized by the complexity and variety of supplies. Workers face the daunting task of ensuring that they receive the correct components. For instance, two identical-looking bearings may have subtle differences in specifications that are critical for their compatibility with specific machines. The manual distinction between such items, accompanied by a meticulous check against delivery documents, is necessary but can also be error-prone, leading to mismatches that might disrupt production or delay critical maintenance.

The Solution: LingLong Insight 1.0

The essence of LingLong Insight 1.0 is its sophisticated combination of image and text understanding, integrated with vast industrial knowledge. This state-of-the-art model is designed to assist workers in verifying the accuracy of industrial supplies effectively. By employing advanced visual recognition techniques, LingLong Insight 1.0 captures images of delivery documents as well as the products themselves. It then cross-references these with existing data, flagging any discrepancies before confirmation of receipt occurs. Such proactive verification minimizes the risk of errors, ensuring that only the correct parts are utilized in production.

With a remarkable 4 billion parameters, LingLong Insight 1.0 has been optimized for edge deployment, striking a fine balance between accuracy and processing efficiency. By integrating Intel's state-of-the-art Core Ultra processors, which feature a unique blend of CPU, GPU, and NPU capabilities, the model processes data right at the workstation. This edge computing approach not only reduces latency but also diminishes the reliance on cloud infrastructure, thereby facilitating immediate and responsive actions on the shop floor.

Leveraging ZKH's Rich Industrial Knowledge

ZKH's extensive industrial expertise, built over three decades, plays a vital role in the development of LingLong Insight 1.0. The company boasts a staggering repository of over 27 million product records and approximately 4 billion product parameters. By standardizing and structuring this data with expert insights, ZKH's AI subsidiary, LingLong Intelligence, has transformed raw industrial knowledge into a strategic resource that underpins scalable AI applications tailored to industrial environments.

Yang Liu, the Chief Technology Officer of ZKH Group and CEO of LingLong Intelligence, emphasized the critical importance of grounded knowledge in the efficacy of industrial AI, stating, "Industrial AI is only as good as the knowledge behind it. Our deep understanding of customer needs enables us to create AI solutions that provide real-world value. Collaborating with Intel allows us to take this intelligence directly to the factory floor, where timely decisions are crucial."

Future Endeavors and Expansion Plans

The launch of LingLong Insight 1.0 represents just the beginning of a larger partnership between LingLong Intelligence and Intel. Together, they are poised to enhance their collaboration in the field of industrial AI, with aspirations to explore a global market presence. By fusing Intel's robust computing technologies with LingLong's profound industrial insights, they aim to develop scalable AI solutions that cater to various industry needs worldwide.

LingLong Intelligence envisions extending the capabilities of the LingLong Insight model beyond verification during goods receipt. Future applications may include thorough checks for inbound and outbound warehouse operations, material verification on production lines, and support for equipment operation and maintenance. Such advancements will harness AI-driven improvements across multiple dimensions of industrial activities.

Conclusion

The collaboration between ZKH Group Limited and Intel signifies a significant leap in the industrial AI domain, promising efficient operations through advanced verification processes. As they embark on this journey, the prospects of AI in transforming industrial landscapes have never been more promising. With the power of LingLong Insight 1.0, the path forward is not just about better AI; it’s about empowering industries to operate with precision and reliability.