ZKH Group Limited says its LingLong Insight 1.0 model is designed to help receiving staff spot specification mismatches in incoming industrial goods before they disrupt maintenance or production.
ZKH describes itself as a Chinese maintenance, repair and operations (MRO) procurement platform. The company says similar-looking parts can have different specifications, and an undetected mismatch could delay maintenance or interrupt production.
At receiving stations, industrial cameras capture delivery records, packaging and labels. The system identifies the goods and checks their specifications against the paperwork, then flags possible discrepancies with related images for staff to review, ZKH said.
ZKH said its AI subsidiary, LingLong Intelligence, launched the four-billion-parameter multimodal model with Intel. A company release posted Nov. 2 described the model as launched at Intel Connection 2026 in Suzhou. Intel’s event listing gives the conference dates as Sept. 22-23.
The system runs on a LingLong Agent Box with Intel Core Ultra processors that combine CPU, GPU and NPU computing, according to ZKH. The company says local processing can reduce delays and reliance on cloud services.
ZKH cites nearly 30 years of industrial-supplies experience and more than 27 million product records as resources behind the model. Yang Liu, ZKH Group’s chief technology officer and LingLong Intelligence CEO, said, “Industrial AI is only as good as the knowledge behind it.”
The company plans to extend the system to warehouse checks, production-line material verification and equipment operation and maintenance. ZKH also says it intends to explore global market opportunities with Intel.
