Self-Assessment Tool
The TUPLES self-assessment tool is an independent tool that increases awareness of best practices and guidelines on Human-centered AI.
The TUPLES Self-Assessment Tool (SAT) is an easy-to-use diagnostic survey designed to help developers and organizations evaluate how well their AI systems for Planning & Scheduling align with EU guidelines on Trustworthy AI. It focuses on key principles such as effectiveness, ethical alignment, robustness, safety, transparency, explainability, reliability, and user interaction.
Structured around five core areas — Human Agency & Oversight, Robustness & Safety, Transparency & Explainability, Privacy & Data Governance, and Fairness & Impact on Work and Environment — the tool guides users through a structured evaluation, offering tailored recommendations based on their answers. At the end of the process, it provides a trustworthiness score (High, Medium, or Low) along with actionable guidance to address gaps or reinforce strengths, supporting the development of reliable and responsible AI systems.
The TUPLES Self-assessment Tool (“SAT”) aims at supporting developers and organizations in the evaluation of the Trustworthiness aspects of Artificial Intelligence (AI) systems used for Planning & Scheduling operations in work environments.
It has been designed as an easy-to-use diagnostic survey, based on the experience matured in the course of the project, that any solution provider may adopt to evaluate the coherence of a specific method with respect to the key EU guidelines on trustworthy, reliable, robust and safe AI.
The goal is to ensure these AI systems are:
- Effective in their intended purpose
- Aligned with ethical standards and regulatory compliance
- Following best practices in key areas such as robustness, safety, transparency, explainability, reliability and user interaction
The tool is structured into five key sections, in line with EU guidelines and incorporating insights from existing tools like ALTAI:
1. Human Agency and Oversight: Evaluates the AI system's interaction with humans and ensures appropriate human autonomy and oversight.
2. Robustness and Safety: Identifies and mitigates safety risks, addresses environmental uncertainty, maintains system integrity, and monitors accuracy and effectiveness.
3. Transparency and Explainability: Assesses data quality, traceability, and the provision of understandable explanations for AI decisions.
4. Privacy and Data Governance: Ensures compliance with data protection regulations and robust data governance measures.
5. Fairness and Impact on Work and Environment: Examines the AI system’s impact on workers, the workplace, and the broader ecosystem.
Before each section begins, users are provided with a brief introduction explaining its relevance and importance in the context of trustworthy AI. These introductions help users understand how their responses will inform the assessment and the tailored recommendations they will receive.
Based on the responses, the tool generates:
- A score classified as HIGH, MEDIUM, or LOW, indicating the system’s alignment with trustworthiness principles
- A detailed recommendation offering tailored, actionable guidance for improvement or maintenance. These recommendations help users address specific gaps or build on existing strengths, ensuring alignment with EU guidelines for Trustworthy AI.