DC Groundwater discharges

Environmental Protection Agency Rijnmond

Supporting AI tool for assessing discharge notifications

Last change on 9th of May 2025, at 13:03 (CET) | Publication Standard 1.0

• Publication category: Other algorithms
• Impact assessment: Field not filled in.
• Status: In use

General information

Theme

Nature and Environment

Begin date

2025-05

Contact information

[e-mail]

Responsible use

Goal and impact

Supporting AI tool for assessing discharge notifications (groundwater). Is the application complete and does it contain all legally required data? The assessment of discharge notifications by an AI bot uses an LLM and the assessment framework from the Environmental Plan. The final judgement whether a notification is Approved or Rejected is up to the handler.

Considerations

The tool saves time when assessing discharge notifications and makes the work for the handler more enjoyable and easier. A drawback can be overconfidence in the result. Therefore, the tool is only available to trained users with sufficient content to assess the result.

Human intervention

The final judgement on whether a report is approved or rejected is up to a human being (the handler).

Risk management

Risks are:

* Interpretability

* Legal correctness

Mitigating measures:

The tool becomes available only to trained users who are sufficiently versed in the content to be able to assess the result.

Legal basis

Environment Act

Links to legal bases

• Omgevingswet: https://wetten.overheid.nl/jci1.3:c:BWBR0037885&z=2024-01-01&g=2024-01-01
• Besluit activiteiten leefomgeving: https://wetten.overheid.nl/jci1.3:c:BWBR0041330&z=2025-01-01&g=2025-01-01

Operations

Data

The algorithm reviews groundwater discharge notifications for the mandatory data (submission requirements) and tests this against applicable laws and regulations. The algorithm uses the assessment framework and is trained with its own datasets of previously human-assessed reports.