What data for this category looks like
The data sources for this category are customer BOMs, Gerber coordinate files, MES systems, mounting machines, SPI/AOI inspection equipment, and reflow oven temperature records. Daily documents include changeover sheets, stencil and furnace temperature records, first article confirmation sheets, SPI/AOI defect reports, maintenance records, and traceability labels. Data updates trigger per work order and each line change. Some inspection data is synchronized in real time. Fields include part number, reference designator, component lot number, soldering temperature, inspection defect type, and others. These fields must match the native format output by corresponding equipment.
Constraints imposed on HTTP interfaces and external systems
This category has high real-time data requirements. Some SPI/AOI inspection data must be synchronized to the audit process in real time. HTTP interfaces must support high-frequency requests and set reasonable timeout thresholds. The field structures of multiple business document types differ significantly. Interface requests must adapt to different formats such as BOMs, inspection reports, and furnace temperature records. Field mapping rules must be passed via the variables parameter. Data updates trigger each time a line change is made or a work order is created. Interfaces must support precise trigger calls by work order ID and line change number to reduce invalid requests. Traceability labels and maintenance records must be bound to a unique work order identifier. Interface returned data must include a traceable unique identification field.
This document is compatible with FastGPT v4.17.0.
How to configure settings
| Configuration Item | Recommended Setting | Rationale |
|---|---|---|
| Interface timeout period | 60-120 seconds | Adapts to response requirements for real-time inspection data synchronization and work order batch processing |
| Identity verification configuration | Fill in according to the preset keys of the external system | Ensures secure connection between FastGPT and production equipment and MES systems |
| Request parameter passing | Bind work order ID and document type identifier | Adapts to field matching requirements for multiple types of business documents |
| Response data extraction | Extract defect results, verification status, and unique traceability codes | Facilitates subsequent business system connection and data retention |
| Retry trigger rules | 3 times, calibrated based on actual measurements | Addresses temporary offline external equipment or network fluctuations |
| Data format adaptation | Adjust according to the output format of corresponding equipment | Supports multiple types of data including BOMs, SPI reports, and furnace temperature records |
Three common configuration mistakes
- Incorrect field mapping passed via the
variablesparameter, resulting in emptyresponseData. The cause is failure to match unique fields of SMT documents, such as reference designator and component lot number. - Interface timeout period set too short, resulting in 504 status code errors. The cause is failure to account for processing time of batch SPI/AOI report data.
- Interface calls not triggered per work order, resulting in repeated submission of line change data. The cause is failure to bind a unique work order ID as the request identifier.
How to confirm configuration is complete
- Call the test interface, pass simulated work order ID and corresponding document data, check if the response returns the expected
responseDatafield. - Simulate external equipment offline or network fluctuations, check if the interface triggers retransmission according to the configured retry rules.
- Compare fields passed by the external system with fields received by FastGPT, confirm the mapping relationship is correct.
- View interface call logs, confirm each request is bound to a unique work order identifier, with no repeated calls.
The values provided are common starting points and should be measured against the reader's own samples.