Accepted Papers, Demo Papers, Posters and Lightning Talks for VehicleSec 2024

Accepted Papers

Accepted Demo Papers

#2 Demo: Efficient and Timely Revocation of V2X Credentials
Gianluca Scopelliti (Ericsson & KU Leuven), Christoph Baumann (Ericsson), Fritz Alder, Eddy Truyen (KU Leuven), Jan Tobias Mühlberg (Université libre de Bruxelles & KU Leuven)

#3 Demo: CAN Security Hands-On Education Platform
Ayaka Matsushita (TOYOTA MOTOR CORPORATION), Tsuyoshi Toyama (TOYOTA MOTOR CORPORATION), Hisashi Oguma (TOYOTA MOTOR CORPORATION), Takeshi Sugawara (The University of Electro-Communications)
DENSO Best Demo Award Runner-up!

#4 Demo: CARLA-based Adversarial Attack Assessment on Autonomous Vehicles
Zirui Lan (Fraunhofer Singapore), Wei Herng Choong (Fraunhofer AISEC), Ching-Yu Kao (Fraunhofer AISEC), Yi Wang (Continental Automotive Singapore), Michael Kasper (Fraunhofer Singapore), Philip Sperl (Fraunhofer AISEC)

#7 Demo: An Open-Source Hardware-in-the-Loop Testbed for Post-Quantum V2V Security Research
 Geoff Twardokus (Rochester Institute of Technology), Hanif Rahbari (Rochester Institute of Technology)

#8 Demo: One Shot All Kill: Building Optimal Attack on Swarm Drones
Minki Lee (DGIST), GangMin Kim (DGIST), JongHyun Kang (DGIST), Hyunwoo Kim (DGIST), Jangwon Lee (DGIST), Hongjun Choi (DGIST)

#9 Demo: Exploiting Cybersecurity Flaws from the ELD Mandate for Trucks
Jeremy Daily (Colorado State University), Jake Jepson (Colorado State University), Rik Chatterjee (Colorado State University)
DENSO Best Demo Award Winner ($100 cash prize)!

#14 Demo: Does Physical Adversarial Example Really Matter to Autonomous Driving? Towards System-Level Effect of Adversarial Object Evasion Attack
Ningfei Wang (University of California, Irvine), Yunpeng Luo (University of California, Irvine), Takami Sato (University of California, Irvine), Kaidi Xu (Drexel University), Qi Alfred Chen (University of California, Irvine)

#15 Demo: SlowTrack: Increasing the Latency of Camera-based Perception in Autonomous Driving Using Adversarial Examples
Chen Ma (Xi’an Jiaotong University), Ningfei Wang (University of California, Irvine), Qi Alfred Chen (University of California, Irvine), Chao Shen (Xi’an Jiaotong University)

#18 Demo: Towards Practical LiDAR Spoofing Attack against Vehicles Driving at Cruising Speeds
Takami Sato (University of California, Irvine), Ryo Suzuki (Keio University), Yuki Hayakawa (Keio University), Kazuma Ikeda (Keio University), Ozora Sako (Keio University), Rokuto Nagata (Keio University), Alfred Chen (University of California, Irvine), Kentaro Yoshioka (Keio University)

#20 Demo: Adaptive Tuning of a Multi-Channel Attack Template for Timing Interference
Ao Li, Marion Sudvarg (Washington University in St. Louis), Han Liu (Washington University in St. Louis), Zhiyuan Yu (Washington University in St. Louis), Chris Gill (Washington University in St. Louis), Ning Zhang (Washington University at St. Louis)

#25 Demo: Towards Automated Driving Violation Cause Analysis in Scenario-Based Testing for Autonomous Driving Systems
Ziwen Wan (University of California, Irvine), Yuqi Huai (University of California, Irvine), Yuntianyi Chen (University of California, Irvine), Joshua Garcia (University of California, Irvine), Alfred Chen (University of California, Irvine)

Accepted Posters

#6 Poster: EV-Fuzz
Ashwin Nambiar (Purdue University), Z. Berkay Celik (Purdue University), Ryan Gerdes (Virginia Tech), Antonio Bianchi (Purdue University)

#9 Poster: Data Sharing in Autonomous Vehicles: Hyperledger Fabric Platform for Secure and Efficient Sharing
Reem Alhabib (University of York), Poonam Yadav (University of York)

#10 Poster: Sensor-based Vehicle Image Authentication
Zhilin Gao (The State Key Laboratory of Blockchain and Data Security, Zhejiang University & ZJU-Hangzhou Global Scientific and Technological Innovation Center), Haoting Shen (The State Key Laboratory of Blockchain and Data Security, Zhejiang University & ZJU-Hangzhou Global Scientific and Technological Innovation Center), Kui Ren (The State Key Laboratory of Blockchain and Data Security, Zhejiang University & ZJU-Hangzhou Global Scientific and Technological Innovation Center)

#12 Poster: Platform for Innovative use of Vehicle Open Telematics (PIVOT)
David Balenson (USC Information Sciences Institute), Christos Papadopoulos (University of Memphis), Jeremy Daily (Colorado State University)

#16 Poster: Robustness of DRL-Based Autonomous Driving to Adversarial Inputs
Ziling He (Waseda University), Tatsuya Mori (Waseda University)

#17 Poster: Bypassing Physical Invariants-Based Defenses in Autonomous Vehicles
Yinan Zhao (Waseda University), Tatsuya Mori (Waseda University)

#22 Poster: Blinding Lights: Attacking Traffic Sign Recognition Through Adversarial Flickering of Streetlights
Alkim Domeke (Clemson University), Pedram MohajerAnsari (Clemson University), Mert D. Pesé (Clemson University)

#23 Poster: Exploring CAN Ringing for ECU Fingerprinting
Ashton McEntarffer (Clemson University), Kevius Tribble (Clemson University), Linxi Zhang (Central Michigan University), Mert D. Pesé (Clemson University)

#24 Poster: PRIDRIVE: An Advanced Privacy Analysis Tool for Android Automotive 
Bulut Gozubuyuk (Clemson University), Mert D. Pesé (Clemson University)

Accepted Lightning Talks

Secure on-sensor machine learning for automotive applications
Mahesh Chowdhary and Swapnil Sayan Saha, STMicroelectronics Inc.

Abstract: To achieve ultra-low-power (uW level), low latency (uS-mS level), small footprint (< few mm), and private inference at the edge of the edge for time-critical automotive systems, sensor manufacturers now integrate custom processing cores directly within the sensor die. There are a large variety of automotive applications from simple application of detecting vehicle stationary condition to complex C-V2X application that require multi-sensor fusion, and can benefit from such compute capability on sensors. These ultra-low-power (uW level) processing capabilities enable moving some computation directly to the sensor from the different control units in automobiles, allowing on-chip sensor fusion, signal conditioning, and running machine learning models on-sensor. We present the latest generation of smart automotive MEMS sensors that can run decision trees and finite state machines within the sensor. We discuss the hardware architecture of these sensors and their integration in automotive systems. We also illustrate no-code automatic machine learning frameworks that can be used to generate smart and private automotive algorithms for these sensors from raw sensor data, as well as tools and techniques to train and deploy performant machine learning models for automotive applications.

CAN Security Hands-On Education Platform
Ayaka Matsushita, TOYOTA MOTOR CORPORATION

Abstract: While the architecture of a future connected autonomous vehicle becomes increasingly more complex, there is a huge demand for the education and training of security engineers with a controller area network (CAN). Although software simulators (e.g., ICSim) is a cost-effective education platform, hands-on experience by using hardware is still a crucial part in learning CAN security. However, learning by attacking real cars is prohibited for ethical and legal concerns. Arduino has been used as a cheap hardware platform for realizing CAN communication, but it does not provide software for simulating the in-car system, such as CAN IDs and data formats, that are critical in learning automotive security. Hardware/software-in-the-loop simulators (HILS and SILS) can address the issue, but they are professional tools and expensive for teaching. Addressing the issue, we developed TestBed1 satisfying the demand. TestBed is composed of two hardware ECUs that provide users with CAN network traffic similar to that of real cars. TestBed is more easily accessible than HILS/SILS due to the simple architecture, and open-source software and hardware designs. In the demonstration, we show TestBed and its use cases, including our teaching experiences in the capture-the-flag (CTF) competitions and a course at a university.

Navigating the Unseen: Human-Inspired Dynamic Risk Assessment in Rare, Out-of-Distribution, and Accident-Prone Scenarios
Shengkun Cui, University of Illinois Urbana-Champaign

Abstract: The wide adoption of machine-learning models facilitates a breakthrough in autonomous driving performance. However, these data-driven models are susceptible to failure in accident-prone scenarios, as those scenarios are underrepresented (out-of-training-distribution, OOD) in the training dataset due to their rarity. Hence, ensuring the safety of autonomous driving in the presence of rare, OOD, accident-prone scenarios via safe and robust decision-making remains a significant challenge.

A critical step for maintaining safety in OOD scenarios is dynamically assessing the risk during autonomous driving. This talk introduces the safety-threat indicator, a risk assessment technique that quantifies safety risks imposed by other actors, singly and collectively, on the ego actor. Inspired by the human driver’s perception of risk, this new technique assesses risk by measuring the change in available “accident escape routes” for the ego actor due to other actors via reachability analysis and counterfactual reasoning.

Our technique provides more comprehensive risk-assessment coverage for various NHTSA-defined accident-prone scenarios than previous methods that use epistemic uncertainty, planning-distribution divergence, or kinematic safety indicators (e.g., time-to-collision). By introducing the safety-threat indicator and its potential applications (e.g., scenario fuzzing, adversarial attacks, and risk-aware-planning), this talk brings a new perspective to safety risk assessment and management for safer autonomous driving.

Virtual-Real Fused Simulation and Testing Platform for Intelligent Connected Vehicle Security
Kun Yang, Zhejiang University

Abstract: Conventional and emerging security challenges have been gravely threatening the security of intelligent vehicles, the key infrastructure of digital society, and thus posing a threat to national security. Countries around the world have promulgated laws and regulations that require intelligent vehicles to undergo security testing before being allowed to drive on the road. Existing security simulations and tests for intelligent vehicles suffer from poor intelligence, disconnection from real scenarios, difficulty to evaluate how the vulnerabilities in one single physical node affect the entire Internet of Vehicles, etc. To address these problems, we have been building a virtual-real fused simulation and testing platform for intelligent connected vehicle security. The virtual-real fused simulation and testing platform generates virtual test scenarios based on real road scenarios. Real vehicles are mapped to virtual scenes using digital twin technology and a sufficient number of virtual vehicles can be generated on demand. Bi-directional transmission of control commands and communication data is realized between the real vehicle and its mapped node. Virtual vehicle nodes are deployed with self-driving algorithmic models for autonomous control. All nodes support installation of V2X applications and telematics stacks. The platform can provide researchers with highly reproducible V2X communication scenarios for functional verification and security testing of V2X protocols and applications. Attacks against V2X can be visualized as they spread from the victim’s vehicle to the entire network. In addition, self-driving confrontation scenario simulation and TARA analysis can also be implemented with the support of this platform. In summary, the platform can provide full-process assistance for intelligent vehicle security assessment.

Security of Sensor Fusion models under Sensor Physical Attacks
Sri Hrushikesh Varma Bhupathiraju, University of Florida

Abstract: Sensor fusion models in autonomous vehicles (AV) are used to compensate for inaccuracies in individual sensors. However, these models are not designed with security in mind. Recent works have demonstrated that physical attacks such as light injection or electromagnetic interference on a single sensor can compromise sensor fusion models, leading to critical consequences on the driving decision of the AV. This is because fusion models assume the integrity of the sensor data. This assumption is broken under physical attacks, leading to incorrect and unsafe decisions from the fusion model. Addressing the security of fusion models requires a systemic approach involving redundancy, anomaly detection, temporal and contextual scene understanding, and real-time dynamic adjustments to sensor behaviors. As sensor fusion models become increasingly popular, let’s define challenges and strategic measures from the intersection between physics and computer science to safeguard them against sensor attacks.