Combining Evasive and Braking Reactions for Safety Reference Models in Automated Vehicles

Riccardo Donà; Konstantinos Mattas; Biagio Ciuffo

cs.ROSep 24, 2026

Combining Evasive and Braking Reactions for Safety Reference Models in Automated Vehicles

Authors: Riccardo Donà, Konstantinos Mattas, Biagio Ciuffo

Organizations: Joint Research Centre for the European Commission, Ispra, IT

Abstract

Computational models of careful and competent human drivers are essential for scenario-based evaluation of automated driving systems (ADS). However, most existing safety reference models primarily focus on longitudinal braking, neglecting the role of evasive steering in human collision avoidance. This paper proposes a hybrid Fuzzy-Safety Model (FSM-H) that integrates longitudinal mitigation and lateral avoidance within a unified behavioral framework. The braking component is governed by Proactive Fuzzy Safety (PFS) metrics, representing the erosion of longitudinal safety margins, while the steering component is driven by Criticality Fuzzy Safety for lane-change (CFS-LC), capturing lateral conflict severity and maneuver feasibility. A finite-state architecture models the sequential escalation from nominal driving to braking and, when necessary, to evasive steering, incorporating perception-reaction time and lane-check delays to reflect human decision processes. The model is evaluated in reconstructed high-criticality cut-in scenarios and compared with braking-only and steering-only reference strategies. Results show that the hybrid approach expands the preventability envelope while maintaining behavioral plausibility and computational tractability. The proposed framework provides a transparent and explainable human reference model suitable for simulation-based ADS safety benchmarking and regulatory assessment.

Figures & tables

Fig. 1: Safety Reference Model (SRM) general formulation

Fig. 2: FSM fuzzy logic. The x FS notation is representative of the different criticality metrics used in this paper (PFS, CFS, …), all of which follow the logic depicted.

Param.	FSM	FSM-LC	FSM-H	Unit
$\tau$	0.75	0.75	0.75	(s)
$\tau_{\mathrm{check\,lane}}$	–	0.5	0.5	(s)
$d_{\mathrm{margin}}$	2.0	2.0	2.0	(m)
$a_{y,\mathrm{comf}}$	–	3.0	3.0	(m/s 2 )
$a_{y,\mathrm{evas}}$	–	5.0	5.0	(m/s 2 )
$b_{\mathrm{ego,comf}}$	4.0	–	4.0	(m/s 2 )

TABLE I: Model parameters for FSM, FSM-LC, and FSM-H.

Fig. 4: FSM vs FSM-LC vs FSM-H cut-in analysis detailed “hard” cut-in scenario breakdown, criticality metrics. The scenario parameters are: $u_{\text{ego}}=$ 130 km/h, $u_{\text{target}}=$ 50 km/h, distance = 50 m, and lateral speed 1 m/s.

Speed (km/h)	Cases	FSM	FSM-LC	FSM-H
70	7200	347	213	258
80	8400	506	287	344
90	9600	713	367	436
100	10800	979	456	538
110	12000	1314	553	650
120	13200	1722	656	771

TABLE II: Crash statistics by initial ego speed.

Fig. 5: Collision kernels density comparison.

Fig. 6: Distribution of impact speeds for the collision scenarios.

Fig. 7: Distribution of TTC when FSM-H switches from braking to steering. In black, the median value.

Explore similar work

PRISM: An Agentic Multi-Model Architecture for Proactive Safety in Autonomous Transportation Systems
Jul 29, 2026Joyjit Roy, Samaresh Kumar Singh, Sushanta DasAgentic Probability
Independent Researcher, IEEE Senior Member, Austin, Texas, USA. · Independent Researcher, IEEE Senior Member, Leander, Texas, USA. · American Center for Mobility, Ypsilanti, Michigan, USA.
Safe2Drive: Evaluating Safe Driving Behaviors of E2E Autonomous Driving Models
May 29, 2026Nishad Sahu, Kalpana Panda, Congyuan Yu +3Human Driving Autonomous Hard Drive Disassembly
1Carnegie Mellon University · 2Birla Institute of Technology and Science Pilani
Modeling Vehicle-Type-Specific Pedestrian Crash Avoidance Behavior in Safety-Critical Interactions Using Smooth-Mamba Deep Reinforcement Learning
May 27, 2026Qingwen Pu, Kun Xie, Hong Yang +2Pedestrian Automated Vehicles
Transportation Informatics Lab, Department of Civil and Environmental Engineering, Old Dominion University, Norfolk, VA 23529, United States · Department of Electrical and Computer Engineering, Old Dominion University, Norfolk, VA, United States · Department of Transportation and Urban Infrastructure Studies, SMARTER Center, Morgan State University, 1700 E Cold Spring Ln, Baltimore, MD 21251, USA

Jul 29, 2026, cs.MAJ/K move · Enter open · S save

PRISM: An Agentic Multi-Model Architecture for Proactive Safety in Autonomous Transportation Systems

Joyjit Roy, Samaresh Kumar Singh, Sushanta Das

Independent Researcher, IEEE Senior Member, Austin, Texas, USA. · Independent Researcher, IEEE Senior Member, Leander, Texas, USA. · American Center for Mobility, Ypsilanti, Michigan, USA.

Agentic Probability

Open paper AskPDF arXiv

Autonomous and intelligent transportation systems operate in complex urban environments where safety depends on interactions among vehicle behavior, environmental conditions, and vulnerable road users (VRUs) such as pedestrians and cyclists. Most advanced driver assistance systems (ADAS) employ reactive mechanisms that activate only after hazards have emerged, a critical limitation underscored by rising VRU fatalities in the United States. This study introduces PRISM (Proactive Risk Intelligence and Safety Management), an agentic multi-model safety architecture that transitions from reactive crash avoidance to proactive, continuous risk management. PRISM employs inverse crash-probability modeling to convert binary crash classifiers into dynamic, interpretable safety scores. Three specialized models addressing trajectory kinematics, environmental risk, and VRU interaction operate concurrently, coordinated by a reasoning layer incorporating reinforcement learning, contextual memory, and feature-level attribution. The system provides graduated safety interventions across four tiers, from silent monitoring to emergency alerts. Unlike rule-based systems with static thresholds, PRISM dynamically adjusts safety parameters in real time. Validated across 1,296 scenarios from three naturalistic driving datasets without dataset-specific retraining, the system yielded a mean safety score of 68 out of 100, classified 77.6% of scenarios as advisory, and flagged a near-miss rate of 3.8%, with 11% of scenarios escalating to intervention or emergency response. Feature attribution consistently identified trajectory risk and VRU proximity as primary safety factors. PRISM provides a unified, interpretable framework for proactive transportation safety with emphasis on VRU risk reduction in dense urban environments.

Abstract

Explore similar work

PRISM: An Agentic Multi-Model Architecture for Proactive Safety in Autonomous Transportation Systems

Safe2Drive: Evaluating Safe Driving Behaviors of E2E Autonomous Driving Models

Modeling Vehicle-Type-Specific Pedestrian Crash Avoidance Behavior in Safety-Critical Interactions Using Smooth-Mamba Deep Reinforcement Learning