Latest Update December '25 - System Builder UX Improvements and Smarter AI with Full System Context

We're excited to announce two major improvements that make CarbonSig more intelligent and easier to use:
(1) Enhanced System Builder UX with improved modals and workflows, and
(2) Expanded AI LCI Suggestions with full System context.

Important action required: 
To see these visual improvements properly in your existing systems, please follow these simple steps:

When opening any system:

• Right-click on blank space in the system canvas
• Select "Arrange" from the menu
• What happens: Your system layout automatically adjusts to display the new visual indicators correctly
• Time required: 2-3 seconds per system

1. Smarter System building with visual data indicators

We're excited to introduce a major enhancement to the System Builder that transforms how you track, verify, and trust your carbon accounting data. The new visual data system makes it instantly clear which data has been reviewed, which needs attention, and what's ready for analysis.

This update brings three powerful improvements that work together to give you complete confidence in your carbon footprint calculations.

What's new

1. Enhanced node information display

2. Three-Color node status system

2. Data review workflow

1. Enhanced node information display

See critical emission data directly on your system canvas—no clicking required.

Nodes now display key emission information directly in Design Mode and Entry Mode, eliminating the need to open each node to see basic data.

What you see on each node type - Input, Direct Emission, and Output nodes:

Display three lines of information:

1. Node name (e.g., "Electricity")
2. Intensity (value + unit) → e.g., "0.0138 kg CO2e/kWh"
3. Total Emissions (value + unit) → e.g., "0.2068 kg CO2e"

Process nodes:

Display three lines of information:

1. Node name (e.g., "Surface Treatment and Anodizing")
2. Total Direct Emissions (value + unit) → e.g., -- because some data is missing from the previous nodes
3. Sum of Scope 1, 2, and 3 emissions (value + unit) → e.g., -- because some data is missing from the previous nodes

Benefits

• Instant overview: Identify high-emission nodes without opening modals
• Quick comparisons: Compare emission intensities across similar inputs
• Faster validation: Spot anomalies or errors at a glance
• Better context: Understand emission flows while building your system
• Enhanced collaboration: Discuss systems with stakeholders using visible data 

2. Three-color node status system

Know exactly what state your data is in—at a glance.

New behavior:

• 🔴 Red = Incomplete (missing required data)
• 🟧 Orange = Draft (has data, needs your review)
• 🟩 Green = Approved (reviewed and confirmed by you)

Understanding the three states

🔴 Red — Incomplete

What it means:

• Node is missing required values
• Required fields not filled (quantity, unit, intensity, etc.)
• Cannot be included in analysis
• Must be completed by you

Behavior:

• Node stays red until ALL mandatory data is entered
• AI-generated values that are missing still show as red
• System cannot run Analyze until all red states are resolved

What to do: → Open the node in Entry Mode → Complete all required fields → Node automatically changes to Orange once data is complete

🟧 Orange — Unverified / AI-Generated / Estimated

What it means:

• Node contains data but hasn't been reviewed by you
• OR contains AI-generated values that is not reviewed
• OR has been edited since last approval
• Can participate in analysis, but results include uncertainty flag

Common reasons for Orange status:

• Data populated by "Build with AI"
• Emission factor suggested by "Suggest LCI with AI" but not reviewed
• You entered data manually but haven't reviewed it yet
• You edited a previously approved (Green) node

Behavior:

• Orange nodes CAN be analyzed (preliminary results available)
• Once you explicitly review and approve → turns Green
• If you edit anything → becomes Orange again until re-approved
• This color merges the old "AI" + "Hybrid" states

What to do: → Open the node in Entry Mode → Review all calculations and data sources → Toggle the approval button ON → Node changes to Green

🟩 Green — Reviewed

What it means:

• All values in the node have been manually reviewed by user
• No unverified AI data
• No missing data
• Ready for full analysis and reporting
• Highest confidence level for carbon calculations

Behavior:

• Only user approval turns a node Green
• If you modify any data, node reverts to Orange
• Indicates you've verified the data is accurate and appropriate

What to do: → Nothing! This node is verified and ready. → Use these nodes confidently for reporting and compliance.

Status Summary Table

3. Data approval workflow

Explicitly confirm your data is accurate—no more ambiguity.

Every node modal in Entry Mode now includes an review toggle button that you must activate to confirm you've reviewed the data.

How it works

Step 1: Enter your data

Click on any node in Entry Mode and complete all required fields:

• Emission source (LCI, CAP, or System)
• Quantity
• Emission category
• Conversion factor

Step 2: Review the Calculations

Check the automatically calculated values:

• Emission intensity (kg CO2e per unit)
• Total emissions (kg CO2e)

Step 3: Review the Data

Toggle the "Review this input" button to ON

Button states:

• Disabled (greyed out): Some data is still missing
• Enabled (white): All required data is complete, ready for your approval

Step 4: Save

Click Assigne → Node changes from Orange to Green

When nodes require Re-review

Nodes automatically revert to 🟧 Orange (draft) status when:

You edit ANY field:

• Change quantity
• Modify conversion factor
• Switch emission source
• Update any calculation parameter

You import a system:

• Even if nodes were Green when exported
• Ensures new user verifies data applies to their context

Why this matters: Any modification could affect carbon calculations. Re-approval ensures you've reviewed the impact of changes.

Review workflow for AI-Generated Systems

When you use "Build with AI" or "Suggest LCI with AI":

All nodes start as 🟧 Orange (AI-Generated/Unverified)

Your review workflow:

Build with AI creates system
       ↓
All nodes show 🟧 Orange
       ↓
Switch to Entry Mode
       ↓
Open each node → Review AI selections
       ↓
Click "Why This LCI?" to understand reasoning
       ↓
Toggle approval ON if satisfied
       ↓
Node turns 🟩 Green
       ↓
Repeat for all nodes
       ↓
System ready for verified analysis

Important: You can run preliminary analysis and CAP generaition on Orange nodes, but Green nodes are required for final reporting, compliance documentation.

How these features work together

Example workflow: Creating a new System

Step 1: Build Structure in Design Mode

• Create processes, inputs, outputs
• All nodes show 🔴 Red (no data yet)
• See node structure clearly

Step 2: Switch to Entry Mode

• Click first Red node
• Enter emission data
• Node turns 🟧 Orange (data entered but not approved)
• Intensity and total emissions now visible on canvas

Step 3: Review and Approve

• Review calculations in modal
• Toggle approval button ON
• Click Accept
• Node turns 🟩 Green (verified)
• Continue for all nodes

Step 4: Visual Confirmation

• Return to Design Mode
• See mix of Red/Orange/Green nodes
• Instantly identify completion status
• All emission values visible on canvas

Step 5: Complete System

• All nodes Green = Highest confidence
• Ready for analysis, reporting, and CAP generation

Special behaviors to understand

Import/Export System behavior

Important change for system sharing:

When you export a system:

• System exports with current node colors (Red/Orange/Green)
• All data and approval states included in export file

When you import that system:

• ALL nodes revert to 🟧 Orange (Unapproved) status
• Even nodes that were Green before export

Why this happens:

• The person importing may be in different context (location, facility, processes)
• Data verification doesn't transfer between users or workspaces
• Each user must confirm data applies to their specific situation
• Ensures data quality and prevents blind acceptance

What to do after importing:

1. Review each Orange node in Entry Mode
2. Verify data matches your context and operations
3. Toggle approval for nodes you've verified
4. Nodes turn Green as you review them

Troubleshooting

Node won't turn green

Problem: I toggled approval but node is still Orange

Possible causes:

• Not all required fields completed
• Approval didn't save (check network connection)
• Page needs refresh

Solution:

1. Verify ALL fields are filled (quantity, unit, emission source)
2. Ensure approval toggle is ON
3. Click Assigne to save
4. Refresh page if needed
5. If issue persists, contact support

Approval Button is Disabled

Problem: Can't toggle the approval button

Why:

• Button remains disabled until ALL required data is complete
• Missing fields prevent approval

Solution:

1. Check for empty fields in the modal
2. Look for red asterisks (*) indicating required fields
3. Complete all mandatory fields
4. Button will enable automatically

Node turned orange after I reviewed it

Problem: My Green node is Orange again

Why this happens:

• You (or a teammate) edited the node data
• Any modification reverts status to Orange
• System requires re-approval after changes

Solution:

1. Review the modified data
2. Verify calculations are still correct
3. Toggle approval button ON again
4. Node returns to Green

All Nodes Orange After Import

Problem: I imported a system and all nodes are Orange, even though they were Green before export

This is expected behavior:

Why:

• Import/export workflow requires re-verification
• Ensures data applies to new user's context
• Prevents blind acceptance of data

Solution:

1. Review each Orange node in your context
2. Verify data applies to your operations
3. Approve nodes that are accurate for you
4. Update nodes that need different values

Can I analyze with orange Nodes?

Yes, but with caveats:

Orange nodes: 

• CAN be included in Analyze calculations 
• CAN generate the CAP
• Provide preliminary carbon footprint estimates
• Results flagged as containing unverified data
• NOT recommended for final reporting or compliance

For final analysis:

• Ensure all nodes are Green (approved)
• Provides highest confidence results
• Suitable for regulatory reporting

Migration: What happened to existing Systems?

Your existing Systems are safe

All existing systems migrated automatically:

✅ Data integrity preserved
✅ Calculations remain accurate
✅ No action required on your part

Status mapping:

Why Orange not Green:

• Previous system couldn't track user verification
• Orange status prompts you to review existing data
• One-time verification process for legacy systems

Recommended action

For critical systems:

1. Open in Entry Mode
2. Review each Orange node
3. Verify data is still accurate and current
4. Toggle approval for verified nodes
5. Node turns Green

For reference-only systems:

• Orange status is fine
• Review and approve when you need to use them
• No urgent action required

2. AI LCI Suggestions Now Understand Your Entire System

The Suggest LCI with AI feature just got significantly smarter — the AI now analyzes your complete system context, not just individual input names.

What Changed

Previously, AI suggestions were based only on:

• Input name
• Input description
• Notes field
• Attached documents

Now, the AI also considers:

• System-level context — your industry, location, and process type
• Connected processes — what happens before and after this input
• Other inputs in the system — materials and energy already defined
• System notes and attachments — uploaded BOMs, EPDs, and specifications
• Process relationships — how inputs flow through your operations

Why this matters

Example scenario: Steel input in Manufacturing system

Result: More accurate, context-aware recommendations that match your specific use case.

Real-World Benefits

Use Case 1: Complex Manufacturing Systems

Situation: Building a laptop computer assembly system

Before:

• AI suggested generic "aluminum" data for laptop chassis
• No differentiation between cast, extruded, or sheet aluminum
• Generic global data didn't match electronics manufacturing

After:

• AI sees: chassis → follows PCB assembly → precedes quality testing
• System context: electronics manufacturing, Asia region
• Other inputs: precision components, cleanroom processes
• Result: Suggests high-grade aluminum alloy (6000 series), sheet form, suitable for CNC machining, appropriate for electronics housing

Time Saved: 15-20 minutes per input Accuracy Improvement: ~40% reduction in manual corrections

Use Case 2: Food & Beverage Production

Situation: Creating a bottled juice production system

Before:

• "Glass bottle" → generic glass data
• No consideration of beverage context
• Missing food-grade requirements

After:

• AI sees: bottles → used in filling process → part of juice packaging
• System context: food manufacturing, Europe, pasteurization process
• Other inputs: fruit concentrate, water, energy for heating
• Result: Suggests food-grade glass bottles, beverage industry standard, appropriate for hot-fill pasteurization

Confidence Score: Improved from Medium to High Verification Time: Reduced by 60%

Use Case 3: Construction Materials

Situation: Modeling concrete production for building project

Before:

• "Cement" → generic Portland cement data
• No regional energy mix consideration

After:

• AI sees: cement → mixed with aggregates → produces ready-mix concrete
• System context: construction materials, Nordic region, cold climate
• Process relationships: batching → mixing → delivery
• Result: Suggests Nordic cement (lower-carbon European production), appropriate for freeze-thaw resistance, regional energy grid accounted for

Accuracy: Climate-specific recommendations ensure realistic carbon footprints

Best practices for maximum benefit

1. Build System structure first

Do this: 

• Complete all processes, inputs, and outputs in Design Mode 
• Connect nodes to show material and energy flows 
• Add descriptions to processes and key inputs 
• Then switch to Entry Mode for data population

Why: AI needs complete structure to understand relationships

2. Use notes & attachments strategically

Do this:

• Attach EPDs and product specifications at system level 
• Add process-specific notes for specialized equipment 
• Include supplier documentation for key materials 
• Upload energy bills for accurate grid data

Why: AI can extract relevant information from all uploaded documents

3. Provide rich System context

Do this:

• Write detailed system descriptions with industry and location 
• Upload complete BOMs and technical specifications 
• Add process notes explaining unique characteristics 
• Specify geographic region and energy sources

Why: More context = smarter AI suggestions

4. Leverage related Inputs

Do this: 

• Populate major inputs first (energy, primary materials) 
• Let AI use these as reference points for related inputs 
• Notice how AI suggestions become more consistent 
• Build up system data iteratively

Why: AI learns from your choices to improve subsequent suggestions

5. Review AI explanations

Do this: 

• Always check "Why This LCI?" for context-based reasoning 
• Verify geographic and process matches 
• Confirm AI understood your system relationships 
• Provide feedback if suggestions miss the mark

Why: Helps you understand AI logic and improve future suggestions

Feedback and support

We value your feedback as we continue to enhance CarbonSig:

Share your experience

Have you noticed:

• Faster system building?
• More relevant AI suggestions?
• Better confidence scores?
• Improved modal usability?

Let us know:

• In-App Feedback: Use the Feedback button to share your experience
• Email Support: support@carbonsig.com for questions or issues
• Help Center: Visit help.carbonsig.app for guides and tutorials

Related documentation

AI features

• Build with AI: AI-Powered System Builder
• Suggest LCI with AI: AI-Powered Emission Data Selection
• How to Write Effective Prompts for Build with AI
• AI LCI Explanation Report

System building

• Quick Start: Build Your First System
• Design Your System (Design Mode)
• Add Data to Your System (Entry Mode)
• Manage Your Systems

Data management

• Reference Emission Data Library
• Latest Update September '25 - Data Marketplace

Keywords

Primary: AI system context, enhanced LCI suggestions, System Builder UX, intelligent carbon accounting, context-aware AI, semantic search carbon data, full system analysis, enhanced modal design

Secondary: process relationship AI, geographic LCI matching, system-wide AI suggestions, improved user experience, carbon footprint automation, smart emission matching, workflow optimization, modal improvements

Technical: system graph analysis, context vector embeddings, multi-factor AI ranking, process topology analysis, Material-UI components, WebGL canvas rendering, semantic LCI search, relationship-aware suggestions

Thank you for being part of the CarbonSig community. Together, we're making carbon accounting more intelligent, efficient, and accurate.

Last updated: December 2025