How it works

Input data

To calculate the impact for your products, we use production data from each step in your supply chain:

Materials:

• Material type and composition (e.g., "recycled polyester", "organic cotton", "nylon")

• Quantity per production step (kg)

• Origin location if materials are transported

Transportation:

• Mode (road, sea, air, rail)

• Distance is calculated automatically from factory locations

Production step:

• Process type (yarn spinning, fabric weaving, dyeing, cut & sew, etc.)

• Output quantity (kg)

• Factory location (we use coordinates for accuracy)

• Production year

Future improvements

This beta version uses the minimum viable data to produce useful results. In future releases, we will add support for additional data sources to improve accuracy:

Energy and utilities:

• Direct uploads from electricity bills (kWh consumption with timestamps)

• Water consumption and wastewater treatment

• Fuel usage for on-site generators or heating

Process-specific data:

• Machine specifications and utilisation rates

• Production batch sizes and waste percentages

• Chemical inputs for dyeing, finishing, and treatments

Our goal is to make it so that the more granular data a producer provides, the more accurate the impact calculation can become.

But you don't need any of this to get started - the current system works with the basic production records that all AWARE member factories have.

How calculations work

Here's what happens when a user submits their production data:

1. Formatting

Your production information is structured into a request for Climatiq's Product Carbon Footprint (PCF) API. Each production step becomes a separate calculation.

Example request for yarn spinning:

{  "product": "Recycled polyester yarn",  "location": "33.9617,118.3379",  "weight": {"weight": 1080, "weight_unit": "kg"},  "manufacturing": [{  "type": "electricity",  "amount": {"energy": 2500, "energy_unit": "kWh"}  }],  "components": [{  "product": "Polyester fiber recycled",  "weight": {"weight": 1080, "weight_unit": "kg"},  "location": "33.9617,118.3379"  }] }

2. Emission factor matching

Climatiq's "Autopilot" system then uses their AI to match materials to emission factors in their database. This happens automatically - we have no direct control over which factor is selected.

For example:

• "Polyester fibre recycled" → Matches to "Plastics (recycled)" at 1.16 kg CO2e/kg

• "Organic cotton fabric" → Matches to "Cotton fabric" at specific emission rate

The system searches across multiple public databases (Bafa, CAEP, BEIS, etc.) and selects the best semantic match for your description.

3. Component calculations

The API calculates emissions for three categories:

Materials: Based on the matched emission factor × quantity

• Example: 1,080 kg recycled PET × 1.16 kg CO2e/kg = 1,252.8 kg CO2e

Transport: Based on distance (from coordinates), mode, and weight

• Example: 461 km road transport × 0.185 tonnes = 13.5 kg CO2e

Manufacturing: Based on electricity consumption × grid emission factor

• Example: 600 kWh × 0.71 kg CO2e/kWh (China grid) = 426.4 kg CO2e

4. Multi-step correction

When you have to calculate chained production steps (yarn → fabric → garment) - maybe because part of your supply chain is missing from the platform or they haven’t enabled impact calculation in their account - the API recalculates each material from scratch.

5. Results aggregation

Final emissions are summed across all production steps and presented with breakdowns by category, production step, and per-unit intensity.

Emission factor matching

This is the most critical- and most uncertain - part of the calculation. Understanding how it works helps you evaluate whether the results make sense. If they do not make sense please reach out, we want to know so we can improve the matching for everyone. 

How Autopilot works

Climatiq's Autopilot uses natural language processing to match material descriptions to emission factors:

What it does:

• Analyses the semantic meaning of your product description

• Searches emission factor databases for similar materials

• Scores potential matches and selects the highest-ranked option

• Assigns a quality label: "accept" (high confidence) or "review" (needs validation)

The only input you have is the product description text itself. However, if you have specific databases or factors you want to use, let us know and we’ll add them to the system. 

Example matching results

Here's what Autopilot selected during our testing with a recycled polyester backpack:

What we're testing

Does Autopilot pick the right factors for your materials?

• Are the emission values reasonable compared to your understanding of these materials?

• Is recycled content being recognised and calculated differently from virgin materials?

• When generic factors are used (like "Textiles" at 7.85 kg CO2e/kg), does that feel like a reasonable approximation?

Which materials cause problems?

• Are there specific materials where Autopilot consistently picks the wrong factor?

• Are there materials we can't calculate at all because no match exists?

• Should we be more specific in how we describe materials? (e.g., "PET flakes recycled" vs "recycled polyester")

Would you prefer manual control?

• Would you rather select emission factors yourself from a list?

• Or is automatic matching acceptable if we show you what was selected and let you flag mismatches?

We need your feedback on this. The emission factors drive the entire calculation, so if Autopilot is picking wrong matches, the results are meaningless.

Please tell us when something looks off.

Multi-step supply chains

When you produce a textile product through multiple steps (fibre → yarn → fabric → garment), a technical challenge emerges: the API recalculates each material from scratch instead of recognising it was already calculated in a previous step.

The problem

Example: Fabric made from yarn

You've already calculated yarn emissions in Step 1:

• 1,080 kg yarn = 1,252.8 kg CO2e (1.16 kg CO2e/kg)

In Step 2, you use 185 kg of that yarn to weave fabric. The API:

1. Matches "yarn" to "Textiles (generic)" emission factor at 7.85 kg CO2e/kg

2. Calculates: 185 kg × 7.85 = 1,452 kg CO2e

3. Returns total fabric emissions including this recalculated yarn

But the yarn was already calculated at 1.16 kg CO2e/kg, not 7.85 kg CO2e/kg. The API has:

• Used a completely different emission factor (generic textiles vs recycled plastics)

• Counted the yarn's emissions twice (once in Step 1, again in Step 2)

Our solution

We automatically detect and correct this double-counting:

Step-by-step correction:

1. Call the API for each production step separately

2. Extract the transport and manufacturing emissions (these are accurate)

3. Subtract the API's material calculation (which is wrong for chained steps)

4. Add back the correct material emissions from the previous step

Fabric example corrected:

API returned: 1,892.1 kg CO2e Subtract wrong yarn calc: -1,452.25 kg CO2e Add correct yarn (Step 1): +214.6 kg CO2e Corrected total: 654.4 kg CO2e

Breakdown of corrected fabric emissions:

• Materials (yarn from Step 1): 214.6 kg CO2e

• Transport (Suqian → Wujiang): 13.5 kg CO2e

• Manufacturing (weaving): 426.4 kg CO2e

• Total: 654.4 kg CO2e ✓

Without this correction, your fabric would show 1,892 kg CO2e - nearly 3x higher than reality.

Data quality and limitations

Understanding what's reliable vs uncertain helps to interpret results correctly.

High confidence

Transport calculations:

• Uses real GPS coordinates and calculates actual distances

• Transport modes (road, sea, air, rail) have well-established emission factors

• Minimal assumptions required

Manufacturing electricity:

• Grid emission factors are published by national authorities

• China: 0.71 kg CO2e/kWh (Climate Transparency, 2021)

• Calculation is straightforward: kWh × grid factor

Medium confidence

Common material emission factors:

• Cotton, polyester, nylon, wool have multiple database entries

• Factors vary by source but are generally consistent (±20%)

• Geographic differences exist but are often minor

Generic textile factors:

• When specific materials unavailable, generic "Textiles" used at 7.85 kg CO2e/kg

• Reasonable approximation for mixed/unknown materials

• Less accurate for materials with extreme values (very high or very low carbon)

Low confidence / needs validation

Recycled material factors:

• Database coverage is limited - many recycled materials fall back to generic factors

• "Plastics (recycled)" factor exists at 1.16 kg CO2e/kg but may not represent all recycled textiles

• Recycled content percentages often not distinguished (30% vs 100% recycled)

Manufacturing energy benchmarks:

• When actual kWh unavailable, we use industry benchmarks:

  • Yarn spinning: 2-3 kWh/kg

  • Fabric weaving: 3-5 kWh/kg

  • Cut & sew: 0.5-1 kWh/kg

• Real consumption can vary significantly based on equipment and efficiency

Trim and accessory materials:

• Often described vaguely ("trims", "accessories", "findings")

• Usually matched to generic textile factors

• Actual carbon footprint depends on specific materials (metal zippers vs plastic buttons)

Known limitation

Limited recycled material database coverage:

• Many recycled textiles match to virgin equivalents

• Recycled cotton, recycled nylon, recycled wool often unavailable

• Falls back to generic factors that may overestimate emissions

What we're testing

This is a beta feature. We need your real-world production data to validate whether this approach works well.

Priority feedback areas

1. Emission factor accuracy

• Do the carbon values match your expectations for your materials?

• Which materials consistently show emission factors that seem wrong?

• Are there industry benchmarks or certifications we should compare against?

2. Missing materials

• Which materials can't be calculated because no match exists?

• Are there textile-specific databases or factors we should integrate?

• Should we build custom emission factors for common materials not in Climatiq's database?

3. Results presentation

• Is the materials/transport/manufacturing breakdown useful?

• What other breakdowns would help? (by production step, by facility, by material type)

• Should we show comparisons? (your product vs industry average, vs previous batches)

How to provide feedback

When something looks wrong or doesn't make sense:

1. Note the specific production step and material

2. Tell us what emission factor was selected (shown in results)

3. Tell us what you expected or what seems more accurate

4. Share any industry data or certifications that contradict the result

You can get in touch through support@wearaware.co or by commenting on the roadmap item.

Technical reference

API and methodology

• API: Climatiq Product Carbon Footprint (PCF) v1-preview2

• Endpoint: https://preview.api.climatiq.io/pcf/v1-preview2/estimate

• Calculation method: IPCC AR4/AR6 GWP100 (depends on emission factor source)

• Scope: Cradle-to-gate (raw materials through production, excludes consumer use and end-of-life)

Emission factor databases

The API searches across multiple sources:

• Bafa: German Federal Office database (European focus)

• CAEP: China Products Carbon Footprint Factors Database

• BEIS: UK Government greenhouse gas reporting factors

• Climate Transparency: Grid electricity emission factors by country

• WRAP, Agribalyse, Oekobaudat: Specialized regional databases

If we’re missing one you want to use, let us know and we’ll get it added.