Periodic Reporting for period 1 - SUperSAFE (SUrrogate measures for SAFE autonomous and connected mobility)
Okres sprawozdawczy: 2023-01-01 do 2025-06-30
The SUperSAFE project—SUrrogate measures for SAFE autonomous and connected mobility—addresses road safety challenges in environments where Connected and Automated Vehicles (CAVs) interact with conventional traffic. Traditional crash-based evaluations limit proactive risk mitigation, necessitating a shift to real-time surrogate measures of safety (SMoS). SUperSAFE pioneers an assumption-free safety assessment framework integrating real-world traffic observations, statistical modeling, and driving simulations.
A key goal is to define critical events and introduce novel SMoS concepts based on proximity and energy dynamics, ensuring a universal safety metric applicable in mixed-traffic and fully automated settings. The project integrates Extreme Value Theory (EVT) with surrogate safety indicators, predicting accident probabilities based on observed interactions rather than crash history. Validation occurs through real-world data collection and controlled simulation experiments.
To enhance transferability, SUperSAFE tests its methodologies across varied environments, leveraging collaborations with institutions in Sweden, Canada, South Korea, and the USA. Beyond academia, the project delivers practical tools for policymakers, road safety professionals, and industry stakeholders, engaging directly with car manufacturers, transport authorities, and technology developers. By shifting from reactive crash analysis to predictive risk assessment, SUperSAFE aligns with Vision Zero and the Safe System Approach, contributing to the safe deployment of CAVs. Its methodologies lay the groundwork for potential commercialization through an ERC Proof of Concept proposal.
A key goal is to define critical events and introduce novel SMoS concepts based on proximity and energy dynamics, ensuring a universal safety metric applicable in mixed-traffic and fully automated settings. The project integrates Extreme Value Theory (EVT) with surrogate safety indicators, predicting accident probabilities based on observed interactions rather than crash history. Validation occurs through real-world data collection and controlled simulation experiments.
To enhance transferability, SUperSAFE tests its methodologies across varied environments, leveraging collaborations with institutions in Sweden, Canada, South Korea, and the USA. Beyond academia, the project delivers practical tools for policymakers, road safety professionals, and industry stakeholders, engaging directly with car manufacturers, transport authorities, and technology developers. By shifting from reactive crash analysis to predictive risk assessment, SUperSAFE aligns with Vision Zero and the Safe System Approach, contributing to the safe deployment of CAVs. Its methodologies lay the groundwork for potential commercialization through an ERC Proof of Concept proposal.
SUperSAFE has advanced proactive risk assessment through three key areas: safety-critical use case identification, new SMoS development, and EVT integration for accident prediction.
A major milestone was the selection of urban and rural test locations with safety-relevant interactions, identified through a workshop with industry, government, and academic stakeholders. Drone and stationary camera monitoring enabled trajectory extraction and behavior analysis.
SUperSAFE introduced a novel, assumption-free SMoS framework defining conflicts based on spatial proximity reduction and energy dynamics, applicable across diverse road environments. The methodology was validated using extracted trajectory datasets, demonstrating robustness before applying it to primary data.
EVT integration with surrogate indicators facilitated a predictive accident risk framework, advancing beyond traditional crash-based models. This work led to two high-impact scientific publications, with a third in progress. Results indicate EVT-based models effectively predict crash probabilities in mixed-mobility settings, supporting the Safe System Approach and Vision Zero.
The driving simulation component progressed with the acquisition of a state-of-the-art simulator at Lund University. Development of a bicycle simulator extends research to vulnerable road users, enabling co-simulation of cyclists and motorized vehicles to study multimodal traffic interactions. Scenario transfer between simulators broadens research opportunities in safety-critical events.
Overall, SUperSAFE has laid a new theoretical foundation for SMoS, integrated EVT for crash prediction, and developed experimental frameworks for validation, positioning itself as a leading initiative in surrogate-based traffic safety research.
A major milestone was the selection of urban and rural test locations with safety-relevant interactions, identified through a workshop with industry, government, and academic stakeholders. Drone and stationary camera monitoring enabled trajectory extraction and behavior analysis.
SUperSAFE introduced a novel, assumption-free SMoS framework defining conflicts based on spatial proximity reduction and energy dynamics, applicable across diverse road environments. The methodology was validated using extracted trajectory datasets, demonstrating robustness before applying it to primary data.
EVT integration with surrogate indicators facilitated a predictive accident risk framework, advancing beyond traditional crash-based models. This work led to two high-impact scientific publications, with a third in progress. Results indicate EVT-based models effectively predict crash probabilities in mixed-mobility settings, supporting the Safe System Approach and Vision Zero.
The driving simulation component progressed with the acquisition of a state-of-the-art simulator at Lund University. Development of a bicycle simulator extends research to vulnerable road users, enabling co-simulation of cyclists and motorized vehicles to study multimodal traffic interactions. Scenario transfer between simulators broadens research opportunities in safety-critical events.
Overall, SUperSAFE has laid a new theoretical foundation for SMoS, integrated EVT for crash prediction, and developed experimental frameworks for validation, positioning itself as a leading initiative in surrogate-based traffic safety research.
SUperSAFE has introduced groundbreaking advancements in SMoS development, predictive risk assessment, and real-world validation of CAV and conventional traffic interactions. The project’s assumption-free SMoS framework eliminates the limitations of traditional conflict indicators reliant on predictive behavior models. Instead, SUperSAFE defines safety-critical interactions based on spatial proximity and energy dynamics, making its methodology universally applicable across various transport environments.
The integration of EVT with surrogate indicators enables real-time crash probability estimation, offering a proactive alternative to historical accident-based assessments. This significantly enhances risk assessment capabilities, aligning with Vision Zero and the Safe System Approach.
Methodologically, SUperSAFE has advanced traffic simulation techniques through a driving simulator framework at Lund University, enabling real-world traffic interactions to be replicated in controlled environments. The integration of a bike simulator broadens applicability to vulnerable road users, pioneering co-simulation of drivers, cyclists, and pedestrians for a comprehensive analysis of multimodal interactions.
Beyond academic contributions, SUperSAFE holds strong industry, policy, and regulatory potential. Further real-world validation will expand dataset applicability, while advanced machine learning integration could enhance predictive accuracy and adaptability. The methodologies developed have commercialization potential in traffic management, vehicle safety software, and infrastructure design. Collaboration with automotive manufacturers, urban planners, and transport authorities is essential for practical implementation.
The project also presents opportunities for intellectual property protection and standardization, with novel SMoS indicators and EVT-based risk assessment methods forming the foundation for patents and regulatory frameworks. Engagement with EU and international road safety agencies ensures alignment with global safety strategies.
Through advancements in accident prediction, safety evaluation, and simulation-based validation, SUperSAFE establishes a universal, assumption-free SMoS framework, integrates predictive risk modeling techniques, and develops co-simulation approaches for mixed-mobility environments. These contributions lay the foundation for future research, commercialization, and policy impact, ensuring emerging mobility technologies contribute to a safer, more resilient transport system.
The integration of EVT with surrogate indicators enables real-time crash probability estimation, offering a proactive alternative to historical accident-based assessments. This significantly enhances risk assessment capabilities, aligning with Vision Zero and the Safe System Approach.
Methodologically, SUperSAFE has advanced traffic simulation techniques through a driving simulator framework at Lund University, enabling real-world traffic interactions to be replicated in controlled environments. The integration of a bike simulator broadens applicability to vulnerable road users, pioneering co-simulation of drivers, cyclists, and pedestrians for a comprehensive analysis of multimodal interactions.
Beyond academic contributions, SUperSAFE holds strong industry, policy, and regulatory potential. Further real-world validation will expand dataset applicability, while advanced machine learning integration could enhance predictive accuracy and adaptability. The methodologies developed have commercialization potential in traffic management, vehicle safety software, and infrastructure design. Collaboration with automotive manufacturers, urban planners, and transport authorities is essential for practical implementation.
The project also presents opportunities for intellectual property protection and standardization, with novel SMoS indicators and EVT-based risk assessment methods forming the foundation for patents and regulatory frameworks. Engagement with EU and international road safety agencies ensures alignment with global safety strategies.
Through advancements in accident prediction, safety evaluation, and simulation-based validation, SUperSAFE establishes a universal, assumption-free SMoS framework, integrates predictive risk modeling techniques, and develops co-simulation approaches for mixed-mobility environments. These contributions lay the foundation for future research, commercialization, and policy impact, ensuring emerging mobility technologies contribute to a safer, more resilient transport system.