Reference only; verify against the target FastGPT release and deployment documentation.
What This Category of Data Looks Like
Internal data for urban public transit comes from on-board smart terminals, dispatch command systems, ticket settlement systems, yard monitoring equipment, and employee work order systems. Data update rhythms fall into three categories:
- On-board location and passenger flow statistics are updated in real time at the second level
- Shift schedules and equipment maintenance records are updated daily
- Annual safety training archives and route operation reports are updated monthly
The structure of a single document includes core identification fields and business fields. Core identifiers include work order ID, route number, and yard ID. Business fields cover fault type, reporting time, processing status, passenger trips, and operating mileage. Field units include trips, kilometers, hours, and similar units.
Constraints Imposed on Workflow Orchestration
Real-time second-level data updates require workflow node trigger intervals to match the data refresh rhythm. This prevents data lag caused by overly long polling intervals. Multi-source heterogeneous data formats require pre-configuring data cleaning nodes in the workflow. These nodes unify field naming rules for work orders and passenger flow data exported from different systems. Regional management of routes and yards requires workflows to support permission verification configured by yard and route dimensions. This limits the visible scope of employees in different positions. Data with different update frequencies require corresponding node execution strategies. Real-time data requires immediate processing. Static report data can be configured for scheduled batch processing.
Configuration Settings
| Configuration Item | Recommended Approach | Basis for This Approach |
|---|---|---|
| Variable reference | Bind the variables parameter to receive the externally passed knowledge base ID | Supports dynamically specifying the target knowledge base when calling the API, adapting to business needs for multiple routes and yards |
| Node trigger condition | Trigger by work order reporting time | Matches the real-time reporting rhythm of public transit work orders, avoiding invalid polling occupying system resources |
| Timeout setting | 600 seconds | Adapts to possible network fluctuations at transit yards, reserving sufficient node response time |
| Result extraction rule | Extract work order ID, route number, yard ID fields | Matches the standard structure of public transit work order documents, ensuring field availability for subsequent processes |
| Retry count | 3 times | Addresses occasional data loss or network jitter when on-board terminals report data |
| Rich text rendering | Use built-in components to insert audio and video markers | Complies with system rendering rules, ensuring normal display of audio and video content |
Three Common Mistakes
- Manually entering original
videooraudiotags results in no audio or video rendering effect in workflow nodes. Reason: FastGPT workflow rich text nodes require inserting audio and video markers via the built-in editing component. Directly entering original tags will not trigger the system's rendering logic. - Passing the knowledge base ID via the
variablesparameter when configuring the knowledge base search node does not take effect. Reason: The receiving rule for this parameter was not declared in advance in the workflow's global variable configuration. This causes the incoming identification information to fail to be parsed correctly during calls. - The text content extraction node returns empty fields or prompts a format error. Reason: The field name used in the extraction rule does not match the field name of the actual data source, such as mistakenly writing "route number" as another non-standard name.
How to Confirm the Configuration Is Correct
- Initiate an API call, pass the
variablesparameter containing the target knowledge base ID, and check whether the workflow correctly calls the specified knowledge base for content retrieval. - Simulate a public transit work order report to trigger the workflow node, view the workflow execution log, and confirm that the extracted fields such as work order ID and route number match the test data.
- Simulate a network fluctuation scenario, trigger a node timeout, and verify whether the system completes the response within 600 seconds and executes the preset retry logic.
- Use the built-in rich text editor to insert audio and video markers, preview the output content of the workflow node, and confirm that audio and video can be displayed normally.
The values provided are common starting points and should be evaluated against the reader's own samples.
Question material comes from public community discussions. Configuration values are common starting points and should be measured against your own samples. Verified on 2026-09-28.