Proportione · Research · Aviaria LineUpdated October 19, 2026 · ← All lines

Avian Line

Responsible AI for high-stakes decisions where mistakes are costly

Summary. In airport operations or any critical decision, AI cannot decide alone. Aviaria investigates how AI expands options and analysis while the person retains authority and responsibility. We begin with bird strike risk in civil aviation, within the EU AI Act and ISO/IEC 42001 framework.

Keywords: Responsible AI · operational safety · bird strike risk · EU AI Act · ISO/IEC 42001

international communication
1
product in development
1
reference regulatory frameworks
3
Proportione booth at REAIM 2026 (A Coruña) and the BirdRisk app in development.

1What the line is

The 20-60-20 framework applied to assisted decision-making

The starting point is bird strike risk: collisions between aircraft and birds are a classic operational safety problem, well-defined, measurable, and extremely demanding.

It is the ideal terrain to test a responsible AI framework that is later applied to other critical decisions: AI provides volume and analysis; the person validates, adjusts, and assumes the decision.

Figure 1 · Assisted decision-making with the 20-60-20 framework, within the guidelines of the EU HLEG, the EU AI Act, and ISO/IEC 42001. Source: own elaboration, Proportione.

2Findings

What we have learned

  1. H1AI never closes the loop alone

    Every output passes through human judgment before becoming a decision.

  2. H2Every step leaves a trail

    The decision is auditable: it is known what the machine proposed and what the person decided.

  3. H3Transferable to other critical decisions

    The role distribution tested in bird risk is applied to other operational safety decisions.

3Product in development

BirdRisk: bird risk in the palm of your hand

iOS operational support application, in development, which displays estimated bird risk by zone and time slot. It has a public landing page and waiting list in Spanish and Portuguese.

BirdRisk app screenBirdRisk app screenBirdRisk app screenBirdRisk app screen

4Publications

The work, as published

First page: Responsible AI decision copilot for aviation risk: the 20-60-20 framework

REAIM 2026A Coruña · ES · ES

Communication

Responsible AI decision co-pilot for aeronautical risk: the 20-60-20 framework

Javier Cuervo

5Full record

All Aviaria publications

2026

  1. Communication · Aviaria

    Responsible AI decision co-pilot for aeronautical risk: the 20-60-20 framework

    REAIM 2026 · A Coruña · ES · Feb 2026 · Javier Cuervo

Full record at ORCID 0009-0006-7134-2894. Works under review are not listed until a decision is made.

6Who signs

Authors

Javier Cuervo

Javier Cuervo

Co-founder and CTO of Proportione · Ph.D. candidate in Business Innovation, Universidade de Aveiro (DEGEIT) · Professor at UNIE

Do you make decisions where security allows no room for error?

We explore with you how responsible AI can strengthen your team's judgment.