Orphaned risk_engine.py appears unused in production and duplicates active risk logic

[Manvikamboz]

Potential orphaned risk engine implementation

While reviewing the risk-analysis code paths, I noticed that backend/risk_engine.py appears to be disconnected from the current production implementation.

Findings

1. No production references

I searched the repository for references to risk_engine and found only a single import in:

• backend/tests/test_risk_engine.py

I found no references in:

• backend/app.py
• backend/conjunction_service.py
• backend/analytics_service.py
• backend/tle_parser.py
• GitHub Actions workflows
• Docker / Render deployment configuration
• README or CHANGELOG documentation

2. Duplicated helper implementations

Several utility functions appear to duplicate implementations already present in backend/app.py, including:

• safe_float
• get_first

This means similar logic currently exists in multiple locations.

3. Separate production risk pipelines

The logic in risk_engine.py appears independent from the active production paths:

• Individual object risk scoring is performed through calculate_risk_score(...) in app.py.
• Conjunction screening is performed through conjunction_service.py.

As a result, modifying risk_engine.py does not appear to affect production behavior.

Potential Impact

The presence of a standalone risk engine that is only referenced by tests may create confusion for future contributors, who may reasonably expect changes to this module to affect live risk calculations.

Question

Is risk_engine.py intentionally retained for future work, or has it become orphaned after the risk and conjunction logic was migrated elsewhere?

If it is no longer part of the active architecture, it may be worth consolidating or removing to reduce maintenance overhead and contributor confusion.

[iTulsi]

Thanks, Manvi, for the detailed investigation. You are correct that the current risk_engine.py is disconnected from the active production pipeline and duplicates some existing logic.

However, OrbitOPS is still under development, and a centralized risk engine is planned for Collision Engine V2. Please do not remove the file for now.

The planned direction is to:

• Keep conjunction_service.py responsible for conjunction detection
• Refactor risk_engine.py into a centralized scoring layer
• Score validated events using miss distance, relative velocity, time to closest approach, object criticality and data confidence
• Remove duplicated helpers and move risk logic out of app.py

I’ll keep this issue open as an architectural refactoring task. Thanks for identifying the inconsistency.

[Manvikamboz]

Thanks for clearing that! @iTulsi
I am with you in this development phrase! Go on!