LCA Data Engine
Intelligent mapping resolves thousands of materials, processes, and energy inputs to their verified scientific source — with built-in validation guardrails. Every value traceable, every result audit-ready.
Always the most accurate data available
Every material is looked up across multiple specialized databases. If one source doesn't have the data, the next one covers it — no unnecessary generic estimates.
Proprietary database with emission factors for INCI ingredients verified against published papers (Shah 2016, Schowanek 2018, Martz 2023).
French ADEME database with processed food product EFs. Queried before Ecoinvent for food because it has specific data (pasta, sauces, oils) instead of generic raw materials.
The global gold standard. Intelligent search with name normalization, polymer resolution, word-order variants, and geographic prioritization (country → region → global).
UK Government emission factors. Covers recycled materials (aluminium, steel, glass) with rigorous data from IAI, World Steel, and FEVE.
Example: "organic cotton velour"
- Normalizes → "organic cotton velour"
- Exact Ecoinvent → not found
- Simplifies → "organic cotton"
- Ecoinvent mapping → found
6 layers of automated precision
Beyond the cascade, every analysis benefits from these specialized systems working behind the scenes.
Country-Specific Electricity
Real grid EF for each manufacturing and sales country. No regional averages — real data for Germany (0.32), France (0.06), Spain (0.27).
Smart Packaging Resolution
"Plastic tube" auto-resolves to polypropylene (cosmetics), PET (food), or PP (other). No guesswork — real Ecoinvent data.
Manufacturing Processes
Manufacturing energy (kWh/kg) pre-loaded from Ecoinvent and scientific literature. Blow molding, injection, milling, dyeing — already calculated.
PEF Upstream Transport
Raw material transport to manufacturer calculated with PEFCR Guidance v6.3 scenarios. Intra-EU (~0.03 kg CO2e/kg) and extra-EU (~0.32 kg CO2e/kg).
Anti-Outlier Validation
Every emission factor is validated against expected ranges per material category. Extreme outliers (>5x range) are automatically rejected.
Multi-Material Weight Allocation
For complex products (chair, laptop), each manufacturing process is applied only to the correct material weight — not to the total product weight.
How it works in practice
Pre-fetch
Before the AI analyzes, the system looks up all material and process EFs from your product in the cascade.
Prompt injection
Verified data is injected directly into the AI context. It doesn't guess — it uses real data from the databases.
Backup tool
If a material wasn't in the pre-fetch, the AI can search in real time using the same cascade. Always verified data.
Other AI tools
Use generic training data. No validation. A material can receive a completely incorrect EF with nobody catching it.
Devera
Every EF is looked up in verified databases, validated against expected ranges, and cited with exact activity and source. Full traceability.
Ready to measure your impact?
Get your Product Carbon Footprint in minutes, not months.