What the data for this category looks like
Data sources include raw material quality certificates, heating furnace systems, press collection data, MES, non-destructive testing instruments, and customer forging drawings. Daily generated documents cover forging process specifications, cutting cards, heating furnace records, forging process cards, mold life records, heat treatment entrustment slips, and ultrasonic testing reports. Data updates occur per piece, per batch, or per furnace heat. Fields include forging ratio, fiber direction, furnace batch number, testing parameters, and more. Units include temperature, pressure, batch number, and more.
What constraints these characteristics impose on the workflow orchestration link
Data updates per furnace heat or batch. Workflows must support accurately associating multi-source production and testing data using furnace batch numbers. Multiple document types require unified verification logic. Workflow nodes must adapt to field extraction rules for different documents. Frequently updated production data must avoid repeated processing. Filter rules must be configured for data trigger nodes. Forging quality defects such as folds and cracks must be verified across the entire process. Workflows must connect exception capture nodes for cutting, heating, forming, testing, and other links.
Configuration Settings
These configurations apply to FastGPT v4.17.0.
| Configuration Item | Recommended Setting | Rationale |
|---|---|---|
| Trigger Rule | Start workflow by furnace batch number | Matches the furnace heat update rhythm to accurately link full-process data for a single furnace heat |
| Global Variable Configuration | Bind variables to furnace batch number and forging ratio requirements | Uniformly pass core business parameters across nodes to avoid repeated entry |
| Node Execution Order | Sort by cutting → heating → forming → non-destructive testing | Aligns with actual forging production process logic to ensure correct verification order |
| Exception Retry Settings | Retry failed nodes 2 times, set timeout threshold to 300 seconds | Adapts to fluctuations in production equipment data collection, reduces false-triggered process interruptions |
| File Parameter Input | Enable file links as input parameters, support up to 1000 MB document uploads | Support uploading documents such as ultrasonic testing reports and process specifications as verification basis |
| Node ID Management | Bind nodeId to each process node | Ensure one-to-one correspondence between workflow nodes and actual production processes, facilitating exception troubleshooting |
Three Common Mistakes
- Phenomenon: A parameter missing error pops up when running the workflow, and the interface displays a prompt that
variablesis not bound. Cause: Core parameters such as furnace batch number and forging ratio are not bound to global variables, leading to unavailability of necessary business data in cross-node verification links. - Phenomenon: Ultrasonic testing reports uploaded from simple tasks cannot pass file links to workflow nodes, and the node parameter bar appears empty. Cause: The configuration to enable file links as input parameters is not enabled, so the file upload link is not connected to workflow nodes.
- Phenomenon: The workflow returns "node execution timeout" after execution, and retries still fail. Cause: No reasonable timeout threshold is set, which cannot adapt to fluctuating delays in heating furnace data collection, resulting in forced process interruption.
How to Confirm Proper Configuration
- Trigger test: Manually enter a real furnace batch number to verify whether the workflow automatically links corresponding batch production data and documents.
- Parameter transfer test: Upload an ultrasonic testing report to confirm whether workflow nodes can receive and parse file links.
- Exception trigger test: Simulate a node execution failure scenario to verify whether the retry mechanism is triggered as configured.
- Variable update test: Modify the forging ratio parameter in the global variable to confirm whether subsequent nodes can obtain the updated parameter value.
The values given are common starting points and should be measured against the reader's own samples.