AI-Supported Import: End-to-End Process Overview

Getting Started

Global Functions

Administration

Extended Planning & Analysis (xP&A)

Consolidation & Financial Planning

ESG Reporting

User Roles and User Interface in ESG Reporting

General Functions in ESG Reporting

All Steps at a Glance – ESG Reporting

Creating Report

Working with Duplicated ESG Reports

Mapping Double Materiality Assessment Results

Configuring the Report Questionnaire

Collecting Data for ESG Report

Collecting and Reviewing ESG Data

Assigning Users as Editors or Reviewers

Displaying Aggregated Responses During Data Collection

Filling Out Report Questionnaire

Reviewing Data

Filling Out the Emissions Module

AI-Supported Import: End-to-End Process Overview

Collecting EU Taxonomy Data

Checking Data

Generating ESG Report

Exporting Data

Creating Custom Emission Factors

Reopening an ESG Report

Canceling an ESG Report

Disclosure Management

XBRL Tagger

Lease Accounting

Banking & Cash Management

Data Collection

Tax Compliance & Reporting

Glossary

With the AI-supported import of consumption data, you can upload your existing emission activity data from MS Excel and have Lucanet's AI automatically convert each row into a structured GHG footprint suggestion — complete with the correct emission factor, scope classification, and CO₂ equivalent values (see also Filling Out the Emissions Module).

The following sections provide an overview of the process, how the AI determines certain values to be included in the footprint suggestions, and an explanation of the AI confidence levels displayed.

You upload one MS Excel file at a time. The AI processes the file and presents you with a set of suggested footprints for review. You then review, adjust where necessary, and accept the footprint suggestions from the current file before you can upload the next one. This ensures data quality and gives you full control over every footprint before it is included in your GHG dataset. The AI processes all relevant Excel sheets within a single uploaded file. It reads each Excel sheet independently and skips rows that represent totals or aggregated summaries.

Every suggested footprint is linked back to its exact origin in your Excel file, i.e., you can see both the sheet name and the original row number from your MS Excel file. This makes it straightforward to cross-reference the AI's output with your source data, verify the numbers, and trace any footprint back to where it came from.

Retrieve valid unit types Before doing anything else, the AI fetches the complete list of valid unit types from the emission factor database (e.g., Volume, Distance, Weight, Money, Energy). This is a prerequisite step that ensures every subsequent search uses only valid, recognized measurement categories.

Read and interpret the file content The AI reads each row across all relevant Excel sheets. For each activity it identifies:

The activity type (e.g., fuel combustion, electricity, freight, business travel)

The geographic location (country or region)

The unit of measurement (liters, km, kWh, EUR, etc.)

The quantity, interpreted in the locale-specific number format of the source data

The reporting year — detected from the file content or sheet context

The original Excel row number, which is appended to every row in the processed data for full traceability

The AI also automatically detects the report year and default region from the file. This data is used as a reference for all subsequent emission factor searches: The reporting year determines which factor vintages are valid, and the detected region anchors the geographic search.

Search for emission factors using a 4-phase strategy For each activity, the AI conducts a structured search within our offered databases and your custom emission factors:

Phase 1 — Discovery: A broad search with minimal constraints (activity keywords and unit type, no region filter) to understand what data is available and learn the naming patterns, available regions, and data sources in the database.

Phase 2 — Region grouping: Based on the detected country, the AI builds geographic tiers — for example, for a German activity: exact Germany → German-speaking cluster (DE, AT, CH) → Central Europe → EU-wide → Global.

Phase 3 — Multi-region search: The AI searches across a geographic cluster in a single query, which is more efficient and automatically provides built-in geographic fallbacks.

Phase 4 — Progressive refinement: If results are insufficient, the AI iterates through broader tiers until suitable candidates are found. If too many results are returned (more than 200), the AI adds more specific keywords or switches to a single-region search to reduce ambiguity.

Select the best-matching emission factor When multiple potential factors are found, the AI applies a clear priority order:

Map and structure the footprint The selected emission factor is mapped to:

The correct GHG scope (Scope 1, 2, or 3)

The appropriate category (e.g., Mobile Combustion, Business Travel, Electricity, Purchase of Goods and Services)

The correct parameter structure — including handling of compound unit types (e.g., weight over distance for freight, passengers over distance for flights). For compound types where both values must be present and cannot be assumed, the AI flags the row as incomplete rather than inventing a missing value.

For every suggested footprint, the AI calculates a confidence score that tells you how certain it is about its generated data. This helps you focus your review on the entries that need the most attention. The overall confidence score is calculated from four weighted factors:

Weighted with

40%

Measures how closely your source text matches the emission factor's activity description:

90–100%

— Exact keyword match (e.g.,

diesel

maps directly to a diesel emission factor)

70–85%

— Synonym or related term match (e.g.,

fuel oil

mapped to

diesel

)

50–70%

— Category-level match only (e.g.,

vehicle fuel

mapped to

diesel

)

30–50%

— Inferred from context (e.g.,

fleet costs

used to infer

diesel

)

0–30%

— No clear match; a fallback factor was used

Weighted with

30%

Measures how precisely the emission factor's geography matches your data's location

100%

— Exact country match (e.g., a German office matched to a German emission factor)

80–90%

— Country within a broader region (e.g., Germany matched to a DACH factor), or no region detected but your default region was used

70–80%

— Continent-level match (e.g., Germany matched to an EU/Europe factor)

50–70%

— No regional overlap; a global factor was used as the best available match

20–40%

— No region detected at all; a global average was applied

0–20%

— Geographic mismatch (e.g., a German activity matched to an Australian factor)

Weighted with

15%

Measures how current the emission factor data is relative to your reporting year

100%

— The emission factor year matches your reporting year exactly

80–90%

— The emission factor is 1–5 years older than your reporting year

60–80%

— The emission factor is 6–20 years older

— The emission factor is newer than your reporting year (methodologically invalid; future data cannot be used for historical reporting)

Weighted with

15%

Measures how many potential factors the AI had to choose from (fewer = more confident)

100%

— Only 1 matching factor found (unambiguous)

70%

— 2–10 matching factors (reasonable disambiguation)

50%

— More than 10 matching factors (high ambiguity)

Formula: confidence = (semantic × 0.40) + (region × 0.30) + (year × 0.15) + (specificity × 0.15)

The combined score is displayed as a color-coded indicator for each suggested footprint:

High confidence

Score ≥ 80%

Review recommended

Score = 50–79%

Needs attention

Score < 50%

Important Notes

For compound activity types (e.g., freight measured in weight over distance), the AI requires both values to be available in your source data. If data is incomplete, the AI will flag the row rather than invent a missing value.