AI System for Checking Emissions Inventory Data Quality

The problem

Unit and version errors may alter an inventory in day-to-day operations. When activity data, unit, period, facility, source document and factor version data are scattered across sensor, document, site or team records, the issue may not be identified in time.

This fragmented data delays preparation of the emissions data quality report and forces teams to search for supporting evidence again by hand. Before making a decision, the environmental engineer or sustainability specialist must complete missing records and verify each inconsistency individually.

A common exception is that a restatement or methodology change may create a valid difference between periods. Incomplete or incorrect automation that does not check this context could lead to emissions being reported incorrectly or a valid record being deleted, so no physical or operational action can be taken directly.

What we set out to improve

• Number of data quality alerts confirmed by the specialist
• Proportion of outputs corrected or rejected by the environmental engineer or sustainability specialist
• Time between the source event and the review output

How the system works

1. 01 Define the review scope and the role of the authorised environmental engineer or sustainability specialist
2. 02 Retrieve the inputs: activity data, unit, period, facility, source document and factor version
3. 03 Check the source identifier, date, calibration or version, and data recency
4. 04 Check the source, period and unit
5. 05 Match the factor version with the calculation input
6. 06 Check the exception: a restatement or methodology change may create a valid difference between periods
7. 07 Place low-confidence or inconsistent records in a separate specialist queue
8. 08 Write the source identifier, rule, output and time to the audit trail
9. 09 Have the environmental engineer or sustainability specialist approve, correct, defer or reject the output

Methods we used

• Schema, unit and version checks without issuing a verification conclusion
• Source, format, time and required-field checks for activity data, unit, period, facility, source document and factor version data
• A missing-data warning instead of an estimate when inputs are incomplete or inconsistent
• A separate exception check outside the main rule
• Joint retention of the source, rule result, emissions data quality report and human decision

Where people stay involved

Data and security

Activity data, unit, period, facility, source document and factor version data are accessed with the narrowest permissions required for the task. Commercial, location and employee information is masked where possible, and only the necessary portion of the data is sent to the model rather than the complete raw dataset. Access, outputs and the environmental engineer's or sustainability specialist's decision are recorded; production does not begin until the organisation approves the retention period and data location.

Who this suits

Frequently asked questions