Problem Statement
Eagle relies on surveillance events and reasoning workflows to analyze activities such as restricted-area intrusion, loitering, crowd formation, and object abandonment. However, contributors and developers currently lack a simple way to generate realistic test data for development, debugging, and experimentation.
Testing often requires manually creating event payloads or obtaining real surveillance datasets, which can slow down development and make reproducing scenarios difficult.
Proposed Solution
Eagle relies on surveillance events and reasoning workflows to analyze activities such as restricted-area intrusion, loitering, crowd formation, and object abandonment. However, contributors and developers currently lack a simple way to generate realistic test data for development, debugging, and experimentation.
Testing often requires manually creating event payloads or obtaining real surveillance datasets, which can slow down development and make reproducing scenarios difficult.
Example output:
{
"person_id": 4,
"event_type": "restricted_zone_intrusion",
"timestamp": "2026-05-31T10:20:11Z",
"location": [120, 85]
}
Suggested location:
utils/synthetic_event_generator.py
Affected Component
LLM Reasoning (services/reasoning/llm.py)
Estimated Difficulty
🟡 Intermediate — Requires understanding of one service
Alternatives Considered
- Manual creation of test JSON files
- Time-consuming
- Difficult to scale
- Hard to reproduce complex scenarios
- Using real surveillance datasets
- Privacy concerns
- Larger storage requirements
- More difficult setup process
- Random event generation without predefined scenarios
- Easier to implement
- Produces less realistic and less useful test cases
Additional Context
This feature would help contributors test reasoning pipelines, validate future AI modules, and quickly generate reproducible surveillance scenarios without relying on external datasets.
The generated events can also be used in future notebooks, demonstrations, benchmarking tasks, and automated testing workflows.
Contribution
Checklist
Problem Statement
Eagle relies on surveillance events and reasoning workflows to analyze activities such as restricted-area intrusion, loitering, crowd formation, and object abandonment. However, contributors and developers currently lack a simple way to generate realistic test data for development, debugging, and experimentation.
Testing often requires manually creating event payloads or obtaining real surveillance datasets, which can slow down development and make reproducing scenarios difficult.
Proposed Solution
Eagle relies on surveillance events and reasoning workflows to analyze activities such as restricted-area intrusion, loitering, crowd formation, and object abandonment. However, contributors and developers currently lack a simple way to generate realistic test data for development, debugging, and experimentation.
Testing often requires manually creating event payloads or obtaining real surveillance datasets, which can slow down development and make reproducing scenarios difficult.
Example output:
{
"person_id": 4,
"event_type": "restricted_zone_intrusion",
"timestamp": "2026-05-31T10:20:11Z",
"location": [120, 85]
}
Suggested location:
utils/synthetic_event_generator.py
Affected Component
LLM Reasoning (services/reasoning/llm.py)
Estimated Difficulty
🟡 Intermediate — Requires understanding of one service
Alternatives Considered
Additional Context
This feature would help contributors test reasoning pipelines, validate future AI modules, and quickly generate reproducible surveillance scenarios without relying on external datasets.
The generated events can also be used in future notebooks, demonstrations, benchmarking tasks, and automated testing workflows.
Contribution
Checklist