Rules Management
How to manage contribution and propagation rules in Darwin
Rules Management
Rules in Darwin determine how data points are transformed into impacts and how those impacts propagate through your supply chain hierarchy. Understanding rules is key to accurate assessments.
Understanding Rules
Darwin uses two types of rules:
Contribution Rules
Define how input data points contribute to impact calculations:
| Rule | Description | When to Use |
|---|---|---|
| Add-up | Adds values together | Multiple inputs of same type |
| Replace | Overwrites calculated values | Direct measurements override models |
| Upgrade Quality | Improves data quality score | Better data available |
Propagation Rules
Define how impacts flow through the entity hierarchy:
| Rule | Description | Example |
|---|---|---|
| Allocation | Splits values between children | 50/50 split between divisions |
| Overlap | Handles duplicate data | Same supplier for multiple products |
The Rules Interface
Navigate to Collect > Rules to access the rules management interface.
Resolve Tab
The Resolve tab shows data points that need attention:
- Allocation conflicts: Data needs to be split between entities
- Overlap issues: Duplicate data detected
- Quality upgrades: Better data sources available
Each row shows:
- Datapoint name and value
- Type and model used
- Business unit and parent entity
- Type of rule needed
- Current status
Trace Tab
The Trace tab shows how rules have been applied:
- View the complete data flow
- Track how values propagate
- Identify sources of impacts
Resolution Strategies
Use the Resolution Strategy dropdown to choose how conflicts are resolved:
| Strategy | Description |
|---|---|
| Auto-Suggest | Darwin suggests resolutions (default) |
| Manual | You decide each resolution |
Start with Auto-Suggest to see Darwin's recommendations, then adjust individual rules as needed.
Resolving Conflicts
Identify Conflicts
Look for rows with status indicators:
- Allocation: Needs percentage split
- Overlap: Needs Add-up or Replace decision
Review the Suggestion
Click the status button to see Darwin's suggestion and alternatives.
Choose Resolution
Select the appropriate rule:
- 50/50: Equal split between entities
- Add-up: Sum all values
- Replace: Use one value, ignore others
- Upgrade Quality: Use higher-quality data
Save Changes
Click Save to apply all pending resolutions.
Common Scenarios
Shared Supplier
When the same supplier provides materials to multiple business units:
Problem: The supplier's impact would be double-counted.
Solution: Use Allocation to split the impact based on:
- Revenue share
- Volume share
- Custom percentages
Multiple Data Sources
When you have both modeled data and direct measurements:
Problem: Which value should be used?
Solution:
- Use Replace to override with direct measurements
- Use Upgrade Quality if direct data is more accurate
Upstream Propagation
When supplier data needs to flow up the hierarchy:
Problem: A farm's impact should appear at the product level.
Solution: Darwin automatically propagates using the Replace (Propagation) rule for pressures.
Rule Status Indicators
| Status | Color | Meaning |
|---|---|---|
| Add-up | Green | Values will be summed |
| Replace | Blue | Value will override |
| Allocation | Orange | Needs percentage split |
| Overlap | Yellow | Potential duplicate |
| Upgrade Quality | Purple | Better data available |
Best Practices
- Review Auto-Suggestions: Check Darwin's recommendations before accepting
- Document Decisions: Add comments explaining non-obvious choices
- Check Regularly: Rules may need updating as you add data
- Validate Results: Compare outputs with expected values
- Use Trace: Follow data flow to understand calculations
Changes in Rules are not automatically saved. Always click Save before leaving the page.
Troubleshooting
Too Many Conflicts
If you see many allocation conflicts:
- Check your entity hierarchy structure
- Ensure entities have correct parent relationships
- Verify scope assignments (Scope 1 vs Scope 3)
Unexpected Values
If calculated impacts seem wrong:
- Go to the Trace tab
- Find the affected data point
- Follow the propagation path
- Check each rule applied
Missing Data
If data points don't appear in rules:
- Ensure they have a valid entity assignment
- Check that the model matching is complete
- Verify the contribution rule is set
Last updated today