The BERTHA Project is pleased to present findings from the paper “Bayesian Network Approach to Building an Affective Module for a Driver Behavioural Model,” published on 10 February 2026 and developed by researchers from UVEG and IBV.

This study focuses on modelling mental load and active fatigue—two critical affective states influencing driving performance—using Bayesian Networks (BNs).

Study Approach

The research team designed a BN that captures the conditional dependencies between relevant variables, offering:

• A transparent graphical representation of how affective states emerge
• A probabilistic inference system that estimates driver state from available evidence
• A flexible, interpretable structure aligned with BERTHA’s multi‑module Driver Behavioural Model

Key Findings

• Bayesian Networks provide a robust and interpretable framework for modelling mental load and fatigue.
• The model can infer the likelihood of specific affective states based on physiological and demographic signals.
• The probabilistic nature of BNs allows the model to handle uncertainty, an essential feature when representing complex human behaviour.

Implications for Human‑Centred Automated Driving

This publication strengthens BERTHA’s efforts in affective modelling, enabling automated systems to better interpret and adapt to human emotional and cognitive variability. The findings support advancements in:

• Driver monitoring for safety‑critical contexts
• Adaptive, human‑aware automated‑driving functions
• More realistic behavioural simulations in CCAM research

This work helps bring affective state estimation into the Driver Behavioural Model, moving BERTHA closer to automated mobility that understands drivers and behaves in a safer, more natural and human‑centred way.

Read the article here.

Acknowledgment: Research conducted under the BERTHA project (GA101076360), funded by the European Union. Views expressed are those of the authors and do not necessarily reflect those of the EU or CINEA.