What the data for this category looks like
Semiconductor wafer fabrication bid-related data typically comes from MES systems, equipment FDC modules, defect detection and metrology devices, wafer map files, process recipe libraries, and experimental datasets. Data updates cover every wafer lot transfer, real-time parameter collection, and daily aggregation. Daily associated documents include process flow cards, lot run records, equipment recipes, SPC control charts, exception handling sheets, defect maps, and yield reports. Most fields relate to process parameters, lot numbers, yield data, and equipment operating status.
What constraints these characteristics impose on forms and interactions
Multi-source, high-frequency updated data for this category requires forms and interactions to support real-time pulling of the latest lot process and yield data. This prevents expired information from being used to prepare quotations. Differences in field formats across multiple document types require forms to support custom field mapping and unit conversion configuration. Ultra-long process chains and limited rework windows require selectable interactive components to support hierarchical filtering. This lets users quickly locate corresponding process and module data. Data errors can lead to serious consequences such as wafer lot scrapping. Forms must add required field validation and data source traceability tags to ensure prepared content can be traced.
How to set up the configuration
The following configurations are developed based on FastGPT v4.17.0, and adapted to the unique data and business requirements of semiconductor wafer fabrication scenarios.
| Configuration Item | Recommended Approach | Rationale |
|---|---|---|
| Form field mapping rules | Bind one-to-one according to the field names of data sources such as MES and equipment FDC | Match standardized data formats exported by different systems |
userSelect component hierarchical structure | Configure three levels: process - module - equipment | Adapt to hierarchical data filtering requirements for ultra-long process chains |
formInput data pull latency | 15–30 minutes | Balance data real-time performance and system load |
| Form validation trigger timing | Full validation before submission | Prevent invalid data from entering the quotation preparation process |
| Dropdown option data source | Bind the real-time updated process recipe library | Ensure quotations use the latest equipment and process parameters |
Three common mistakes
- The
userSelectcomponent dropdown list is empty, making it impossible to select the corresponding process module. This occurs because the real-time updated process recipe library data source is not bound, or the data source update frequency configuration exceeds the preset threshold. - The form returns unexpected referenced content after submission, even when the reference switch is turned off. This occurs because global variable dynamic assignment rules are not properly configured, causing leftover context data to be brought into the preparation process.
- Form field validation fails but no clear prompt is displayed. This occurs because custom validation rules are not configured for unique fields such as nanoscale parameters and lot numbers, and only general validation logic is used.
How to confirm the configuration is complete
- Open the configuration interface of the
userSelectcomponent, confirm that the dropdown options include the preset three-level hierarchical structure of process, module, and equipment. - Trigger the data pull action of the
formInputcomponent, check that the pulled fields fully match the actual fields of data sources such as MES and equipment FDC. - Submit a test form containing abnormal parameters, confirm that the system pops up validation prompts for the corresponding fields.
- Turn off the reference switch and submit the form, confirm that the returned content does not include unauthorized external document content.
The values provided here are common starting points and should be measured against the reader's own samples.