Imagine if self-driving cars drove exactly like a cautious human instead of like a rigid robot. This project creates a digital 'brain' that mimics how real people react to traffic and stress. By teaching cars to behave like us, other drivers will trust them more and feel safer sharing the road.
Autonomous vehicles often behave in ways that feel unnatural or unpredictable to humans, leading to low user trust and safety risks in mixed traffic. There is currently no scientifically validated model that allows manufacturers to design and test these human-like interactions digitally.
A probabilistic Driver Behavioural Model (DBM) based on Bayesian Belief Networks, integrated into an open-source HUB and the CARLA simulation platform.
If you are an OEM dealing with low consumer trust in self-driving features — this project developed a Driver Behavioural Model that makes autonomous responses more human-like. This increases user acceptance and safety in mixed traffic.
If you are a Tier 1 supplier dealing with the difficulty of testing autonomous software in real-world chaos — this project developed a simulation tool using CARLA and Bayesian networks. It allows you to validate components digitally before physical production.
If you are a software provider dealing with unrealistic AI drivers in virtual tests — this project developed an open-source HUB of driver archetypes. This provides a scientifically grounded way to build more realistic traffic simulations.
Based on available project data, specific pricing for the end-user is not mentioned, although the model is being implemented in an open-source HUB.
Yes, the project specifically aims to create a scalable and probabilistic model implemented in a HUB to ensure global scalability for industry use.
The project implements the model in an open-source HUB to share it with the scientific community and industry for validation.
The model is designed to be integrated into Electronic Control Units (ECUs) to generate human-like responses in autonomous vehicles.
The project runs from 2023-11-01 to 2026-10-31, moving from TRL 2 to TRL 4.
The consortium is highly industry-oriented with a 41% industry ratio, comprising 17 partners across 7 countries. The presence of 7 industry partners, including 3 SMEs and a coordinator that is an SME, suggests a strong focus on commercial viability and practical application rather than pure academic research.
Contact Instituto de Biomecanica de Valencia for technical specifications on the DBM HUB.
Contact us to find partners for TRL 4 validation of human-centric driving models.
