{"id":{"repo_id":"msu","oai_identifier":"oai:d.lib.msu.edu:etd_52713"},"canonical_url":"https://search.ndltd.org/etd/msu/oai:d.lib.msu.edu:etd_52713","repository":{"repo_id":"msu","name":"Michigan State University","base_url":"https://d.lib.msu.edu/oai"},"display":{"title":"PHYSICAL ATTACKS ON AUTONOMOUS VEHICLES IN SENSOR PERCEPTION AND CHARGING SYSTEM","abstract":"Thesis (Ph.D.)--Michigan State University. Computer Science - Doctor of Philosophy, 2025","abstract_html":"Thesis (Ph.D.)--Michigan State University. Computer Science - Doctor of Philosophy, 2025","abstract_has_math":false,"creators":["Zhou, Ce"],"institution":null,"degree_name":null,"degree_level":null,"degree_discipline":null,"degree_department":null,"school":null,"contributors":["Yan, Qiben","Liu, Sijia","Siegel, Josh","Bopardikar, Shaunak D"],"advisors":[],"committee_chairs":[],"committee_members":[],"year":2025,"date_issued":"2025","date_published":"2025","updated_at":"2026-11-18T18:59:26Z","subjects":["Computer science","Artificial intelligence","Electrical engineering"],"languages":["eng"],"rights":["Attribution-NonCommercial-NoDerivatives 4.0 International"],"rights_urls":[],"identifier_entries":[{"key":"dc:identifier","label":"Identifier","values":["etd:52713","local:Zhou_grad.msu_0128D_21684"],"render_values":[{"text":"etd:52713","href":null,"code":true},{"text":"local:Zhou_grad.msu_0128D_21684","href":null,"code":true}]}],"person_identifier_entries":[]},"links":{"outbound_url":"https://doi.org/doi:10.25335/xpps-ks62","outbound_label":"DOI","outbound_source":"dc:identifier","record_url":"https://doi.org/doi:10.25335/xpps-ks62","record_label":"Repository record","record_kind":"doi","record_source":"dc:identifier","download_url":null,"download_label":null,"download_source":null,"access_status":null,"embargo_until":null,"access_note":null},"source_record":{"url":"https://d.lib.msu.edu/oai?verb=GetRecord&metadataPrefix=oai_dc&identifier=oai%3Ad.lib.msu.edu%3Aetd_52713","prefix":"oai_dc"},"metadata_groups":[{"id":"people","label":"People","entries":[{"key":"dc:contributor","label":"Contributor","values":["Yan, Qiben","Liu, Sijia","Siegel, Josh","Bopardikar, Shaunak D"]},{"key":"dc:creator","label":"Author","values":["Zhou, Ce"]}]},{"id":"academic_context","label":"Academic Context","entries":[{"key":"dc:date","label":"Dc Date","values":["2025"]},{"key":"dc:relation","label":"Dc Relation","values":["Electronic Theses & Dissertations"]},{"key":"dc:type","label":"Dc Type","values":["text","Electronic dissertations"]}]},{"id":"subjects_keywords","label":"Subjects and Keywords","entries":[{"key":"dc:subject","label":"Dc Subject","values":["Computer science","Artificial intelligence","Electrical engineering"]}]},{"id":"language_rights","label":"Language and Rights","entries":[{"key":"dc:language","label":"Dc Language","values":["eng"]},{"key":"dc:rights","label":"Dc Rights","values":["Attribution-NonCommercial-NoDerivatives 4.0 International"]}]},{"id":"identifiers","label":"Identifiers","entries":[{"key":"dc:identifier","label":"Identifier","values":["etd:52713","local:Zhou_grad.msu_0128D_21684","https://doi.org/doi:10.25335/xpps-ks62"]}]},{"id":"additional","label":"Additional Metadata","entries":[{"key":"dc:description","label":"Description","values":["Thesis (Ph.D.)--Michigan State University. Computer Science - Doctor of Philosophy, 2025","Autonomous vehicles are reshaping modern transportation, aiming to improve safety, efficiency, and accessibility. Sensor perception plays a foundational role in autonomous driving and flying, providing the vehicle with a comprehensive understanding of its environment, which is essential for safe and efficient operation in dynamic, unpredictable surroundings. On the other hand, the charging protocol in autonomous vehicles manages critical interactions, such as energy transfer and system monitoring, to ensure the vehicle's safety, reliability, and overall operational integrity. However, in recent research on autonomous vehicles, security vulnerabilities in sensor perception and charging systems have emerged as significant concerns. To comprehensively understand the potential security threats to autonomous vehicles, this dissertation investigates physical attacks that disrupt either autonomous driving or battery charging operations by exploiting vulnerabilities in sensor perception systems and charging protocols. On the perception side, the work focuses primarily on depth estimation and object detection in both autonomous flying and driving scenarios, and introduces three distinct physical attacks targeting these components. On the charging side, the dissertation examines the electric vehicle charging protocol, presenting one physical-layer attack that manipulates the charging process through protocol-level vulnerabilities.First, this dissertation proposes doublestar, a novel attack on stereo depth estimation used for obstacle avoidance in autonomous vehicles and drones. By exploiting stereo matching weaknesses and optical lens flare effects, doublestar injects false depth perceptions using projected light sources, deceiving commercial stereo cameras and drones to cause sudden stops or erratic drifting. %, revealing critical flaws in stereo-based obstacle avoidance. Second, this dissertation introduces LensAttack, a physical attack using concave and convex lenses strategically placed on the vehicle's camera. By exploring the vulnerability of monocular depth estimation algorithms, LensAttack can manipulate depth perception to produce false object distances. It reduces the accuracy of monocular depth estimation significantly, raising concerns about the reliability of vision-based autonomous driving systems. Third, this dissertation presents an adversarial 3D projection attack on object detection in autonomous vehicles. By using dynamic color and geometric transformations on 3D surfaces, the attack can generate projectable adversarial patches on 3D objects to deceive object detection algorithms transiently and flexibly using projectors, potentially leading to traffic accidents. Fourth, this dissertation proposes ChargeX, which aims to manipulate the charging states or charging rates of electric vehicles to disrupt the normal charging schedules. By investigating the communication mechanisms between chargers and electric vehicles, the lack of protection for the authenticity in the SAE J1772 charging control protocol is exposed. ChargeX inserts a hardware attack circuit to strategically modify the charging control signals, causing denial of service or degradation of the battery performance.We have implemented and extensively evaluated the proposed doublestar, LensAttack, 3D projection attack, and ChargeX in both simulation and the physical world. These attacks were rigorously tested across diverse platforms and environments, including drones, self-driving cars, and electric vehicle charging infrastructures. The comprehensive evaluations validate not only the feasibility and practicality of each attack but also their potential impacts on safety-critical operations in autonomous vehicle systems.This dissertation concludes with an in-depth discussion of the limitations of current work and future research directions, highlighting both the challenges and opportunities in securing next-generation autonomous vehicle systems. These include exploring attacks and defenses that span multiple layers, from perception and sensor fusion to protocol validation and hardware-level protection, as well as understanding how the introduction of emerging technologies, such as foundation models, may open new attack surfaces or defense possibilities. By exposing underexplored physical vulnerabilities and developing effective attacks, this dissertation aims to provide not only a security warning but also a foundation for proactive defense design. The insights derived from these attack studies pave the way for future research toward building more robust, adaptive, and resilient autonomous vehicle systems.","Description based on online resource. Title from PDF t.p. (Michigan State University Fedora Repository, viewed ).","Includes bibliographical references."]},{"key":"dc:format","label":"Dc Format","values":["application/pdf","1 online resource (175 p.) : ill."]},{"key":"dc:title","label":"Title","values":["PHYSICAL ATTACKS ON AUTONOMOUS VEHICLES IN SENSOR PERCEPTION AND CHARGING SYSTEM"]}]}],"canonical_facts":{"dc:contributor":["Yan, Qiben","Liu, Sijia","Siegel, Josh","Bopardikar, Shaunak D"],"dc:creator":["Zhou, Ce"],"dc:date":["2025"],"dc:description":["Thesis (Ph.D.)--Michigan State University. Computer Science - Doctor of Philosophy, 2025","Autonomous vehicles are reshaping modern transportation, aiming to improve safety, efficiency, and accessibility. Sensor perception plays a foundational role in autonomous driving and flying, providing the vehicle with a comprehensive understanding of its environment, which is essential for safe and efficient operation in dynamic, unpredictable surroundings. On the other hand, the charging protocol in autonomous vehicles manages critical interactions, such as energy transfer and system monitoring, to ensure the vehicle's safety, reliability, and overall operational integrity. However, in recent research on autonomous vehicles, security vulnerabilities in sensor perception and charging systems have emerged as significant concerns. To comprehensively understand the potential security threats to autonomous vehicles, this dissertation investigates physical attacks that disrupt either autonomous driving or battery charging operations by exploiting vulnerabilities in sensor perception systems and charging protocols. On the perception side, the work focuses primarily on depth estimation and object detection in both autonomous flying and driving scenarios, and introduces three distinct physical attacks targeting these components. On the charging side, the dissertation examines the electric vehicle charging protocol, presenting one physical-layer attack that manipulates the charging process through protocol-level vulnerabilities.First, this dissertation proposes doublestar, a novel attack on stereo depth estimation used for obstacle avoidance in autonomous vehicles and drones. By exploiting stereo matching weaknesses and optical lens flare effects, doublestar injects false depth perceptions using projected light sources, deceiving commercial stereo cameras and drones to cause sudden stops or erratic drifting. %, revealing critical flaws in stereo-based obstacle avoidance. Second, this dissertation introduces LensAttack, a physical attack using concave and convex lenses strategically placed on the vehicle's camera. By exploring the vulnerability of monocular depth estimation algorithms, LensAttack can manipulate depth perception to produce false object distances. It reduces the accuracy of monocular depth estimation significantly, raising concerns about the reliability of vision-based autonomous driving systems. Third, this dissertation presents an adversarial 3D projection attack on object detection in autonomous vehicles. By using dynamic color and geometric transformations on 3D surfaces, the attack can generate projectable adversarial patches on 3D objects to deceive object detection algorithms transiently and flexibly using projectors, potentially leading to traffic accidents. Fourth, this dissertation proposes ChargeX, which aims to manipulate the charging states or charging rates of electric vehicles to disrupt the normal charging schedules. By investigating the communication mechanisms between chargers and electric vehicles, the lack of protection for the authenticity in the SAE J1772 charging control protocol is exposed. ChargeX inserts a hardware attack circuit to strategically modify the charging control signals, causing denial of service or degradation of the battery performance.We have implemented and extensively evaluated the proposed doublestar, LensAttack, 3D projection attack, and ChargeX in both simulation and the physical world. These attacks were rigorously tested across diverse platforms and environments, including drones, self-driving cars, and electric vehicle charging infrastructures. The comprehensive evaluations validate not only the feasibility and practicality of each attack but also their potential impacts on safety-critical operations in autonomous vehicle systems.This dissertation concludes with an in-depth discussion of the limitations of current work and future research directions, highlighting both the challenges and opportunities in securing next-generation autonomous vehicle systems. These include exploring attacks and defenses that span multiple layers, from perception and sensor fusion to protocol validation and hardware-level protection, as well as understanding how the introduction of emerging technologies, such as foundation models, may open new attack surfaces or defense possibilities. By exposing underexplored physical vulnerabilities and developing effective attacks, this dissertation aims to provide not only a security warning but also a foundation for proactive defense design. The insights derived from these attack studies pave the way for future research toward building more robust, adaptive, and resilient autonomous vehicle systems.","Description based on online resource. Title from PDF t.p. (Michigan State University Fedora Repository, viewed ).","Includes bibliographical references."],"dc:format":["application/pdf","1 online resource (175 p.) : ill."],"dc:identifier":["etd:52713","local:Zhou_grad.msu_0128D_21684","https://doi.org/doi:10.25335/xpps-ks62"],"dc:language":["eng"],"dc:relation":["Electronic Theses & Dissertations"],"dc:rights":["Attribution-NonCommercial-NoDerivatives 4.0 International"],"dc:subject":["Computer science","Artificial intelligence","Electrical engineering"],"dc:title":["PHYSICAL ATTACKS ON AUTONOMOUS VEHICLES IN SENSOR PERCEPTION AND CHARGING SYSTEM"],"dc:type":["text","Electronic dissertations"]},"updated_at":"2026-11-18T18:59:26Z"}