From Bill of Materials to PCF in Minutes: How AI Changes the Calculation | Carbmee
How Automatic Emission Factor Matching Works
Calculating a product carbon footprint has traditionally been a slow, resource-intensive process. Mapping every item in a bill of materials to an emission factor, sourcing missing weight data, and reconciling secondary estimates with supplier reality can take weeks for a single product and becomes unmanageable at portfolio scale. Carbmee's AI-powered emission factor matching fundamentally changes this equation.
The starting point for a PCF calculation is a bill of materials: a structured list of every component, material, and input that goes into a finished product. This data can be imported directly from an Excel export out of a PLM or ERP system, and carbmee EIS™ handles column mapping automatically − so even if the supplier column in the source file is labeled "vendor" rather than "supplier," the system resolves the mismatch without manual reformatting. During the import, the data is also validated: missing required fields are highlighted, and built-in outlier detection flags values outside normal ranges.
Once the BOM is imported, the platform automatically assigns an emission factor to every leaf-level item using AI-powered matching against databases including ecoinvent. Where better data already exists in the platform, for example primary data already collected from suppliers, it is used instead. The matching engine draws on item descriptions, material categories, and any other natural language available in the source data. The more descriptive the input data, the higher the confidence of the match.
Each match comes with a confidence score, allowing teams to immediately identify where Carbmee's AI is highly certain and where human review or supplier input would add the most value. Rather than checking thousands of items one by one, teams can filter by confidence level combined with emissions impact − focusing attention on the materials that are both high-emission and low-confidence, where the return on data enrichment is greatest.
Filling the Weight Gap With AI
A common obstacle in moving from spend-based to activity-based emission factors is missing weight data. Many purchasing systems record what was bought and from whom, but not the weight of each item − the input that activity-based calculation requires.
For items where weight is absent, the platform generates an AI-based weight estimate, with a confidence score and a plain-language explanation of how the figure was derived. This enables a complete activity-based baseline to be built even from imperfect source data, and the estimates can later be refined either by internal experts or through supplier collaboration.
Supplier Collaboration: Confirming or Refining the Model
Rather than sending blank data request forms to suppliers and waiting for responses, carbmee EIS™ takes a different approach: pre-calculating an emission model for every relevant SKU and presenting it to the supplier for confirmation or refinement.
Suppliers with access to the platform can see the secondary model that has been built for their products − including the bill of material composition and the energy used − and update it where their real production data differs. For example, if they have already transitioned to renewable energy at a particular site, or if their actual material mix differs from the secondary assumptions. Changes flow back into the PCF automatically, and a full changelog records every modification made, whether by the AI, an internal user, or a supplier.
This approach is more productive than a blank questionnaire because it gives suppliers a concrete starting point rather than asking them to produce data from scratch. It also concentrates engagement where it matters most: hotspot suppliers in scope 3 whose emissions account for the largest share of the product footprint.
Quality Gates and Auditability
As primary data arrives from suppliers, a Carbmee's Quality Gates module allows teams to review and compare the incoming data against the secondary model. Teams can approve supplier data, request revisions, or flag it for further review − all with a documented audit trail.
For organizations preparing for third-party verification, this structured workflow is important. Every data point in the platform carries a provenance record: where it came from, when it was submitted, and what changes have been made to it. Supporting documents such as energy certificates, test reports, or verification opinions can be attached directly to the relevant items, making the audit package straightforward to assemble.
The result is a PCF that is not only calculated quickly inside carbmee EIS™, but is built to be verified.
