Linjiang factory of Advance Gearbox, a key supplier for wind turbine gear shafts in China, struggled with fluctuating orders, inconsistent supplier lead-times, and overstocking at multiple process stages.

Eddso’s team deployed a layered approach to visualize material flow, classify supplier risk tiers, and build a predictive scheduling engine based on real-time consumption. The result: a stable, scalable and digitally visible supply network.

The supply chain was fragmented across machining, heat treatment, and third-party inspection vendors, with minimal coordination and excess lead-time buffers.

• Fluctuating Customer Forecasts
  Late changes from OEM clients caused schedule crashes and order reprioritization.
• Overlapping Inventory Layers
  Materials accumulated at each production stage due to lack of real-time planning.
• Inconsistent Sub-supplier Commitments
  Heat treatment vendors missed slots, causing backlog across the line.
• No Tiered Risk Management
  Supplier reliability was not quantified; all were treated equally in planning.

To strengthen delivery reliability and vendor tiering, a hybrid system integrating risk scoring and TPM routines was deployed.

• Supplier Tier Classification
  Grouped vendors into Red-Yellow-Green tiers based on delivery scores and lead-time adherence.
• TPM-Linked Reliability Checks
  Added daily issue boards + real-time maintenance routines to address supplier-induced line stops.
• Quick-Connect Layouts for Resilience
  Equipment and handling zones were modularized to flexibly absorb disruptions from lower-tier vendors.