BEGIN:VCALENDAR
VERSION:2.0
PRODID:-//132.216.98.100//NONSGML kigkonsult.se iCalcreator 2.20.4//
BEGIN:VEVENT
UID:20260906T072449EDT-4123EuMGXC@132.216.98.100
DTSTAMP:20260906T112449Z
DESCRIPTION:Abstract\n\nIn recent years\, autonomous vehicles (AVs) control
 led by advanced machine learning techniques have significantly gained in p
 opularity. With many promising functionalities\, citizens are quick to tra
 nsition from regular road vehicles to (at least partially) autonomous vehi
 cles for their daily travels. However\, as the number of AVs on our roads 
 increases\, so do the related safety assurance concerns.\n\nTo identify an
 d address such safety concerns\, researchers and practitioners often refer
  to existing safety standards for autonomous (e.g. ISO 21448) and for regu
 lar vehicles (e.g. ISO 26262-1). Of note is the level of detail in such st
 andards: although there exist requirements for the individual components i
 nvolved in AVs\, safety standards place system-level requirements and rest
 rictions on the AV-under-test. Unfortunately\, research has shown that des
 pite being efficient in their scope\, component-level testing approaches o
 ften do not adapt well to the system level. Therefore\, system-level testi
 ng approaches must be derived independently and may make certain assumptio
 ns wrt. the correctness of the various underlying components.\n\nSystem-le
 vel safety assurance approaches for AVs are often based on adaptations of 
 existing approaches for general software systems. However\, existing resea
 rch suggests that such adaptations are not adequate for AV testing. On one
  hand\, upfront design time verification of AVs is practically infeasible 
 considering the potentially infinite number of environments (contexts) an 
 AV must interact with. On the other hand\, runtime techniques\, such as on
 -road monitoring\, are unsustainable as they place untested or partially t
 ested AVs on real roads\, which poses a serious threat to the safety of su
 rrounding vehicles and pedestrians. To address such testing challenges\, r
 ecent safety assurance approaches adopt the scenario-based testing paradig
 m: they test AVs by (1) automatically deriving traffic scenarios\, (2) exe
 cuting them in simulation and (3) evaluating the system-level safety of th
 e AV-under-test.\n\nIn this thesis\, I propose a scenario-based AV testing
  approach that builds upon automated generation of critical test traffic s
 cenarios. I provide contributions in accordance to three foundational rese
 arch questions. As an initial step\, I propose (FRQ1) a multi-faceted\, fo
 rmal scenario specification language that incorporates relevant traffic co
 ncepts at various levels of abstraction for adequate representation of AV 
 test cases. To evaluate the (practical) relevance of traffic scenarios\, I
  propose (FRQ2) various coverage metrics at different abstraction levels. 
 In particular\, such metrics incorporate concepts related to potential dan
 ger in traffic scenarios (e.g. collisions\, near-misses).\n\nTo complete t
 he AV testing workflow\, I then propose (FRQ3) various approaches that der
 ive consistent\, simulation-ready traffic scenarios with high abstract cov
 erage from abstract specifications given as input. I compare different app
 roaches in accordance to the particularities of the input specification la
 nguage it handles and to the relevance of traffic scenarios it derives. As
  a practical outcome of my contributions\, I provide safety evaluation dat
 a and analysis for three state-of-the-art AV controllers (i.e. TransFuser\
 , Dave2 and BeamNG.AI) within two simulation environments (i.e. CARLA\, Be
 amNG.tech).\n
DTSTART:20240823T143000Z
DTEND:20240823T163000Z
LOCATION:Room 603\, McConnell Engineering Building\, CA\, QC\, Montreal\, H
 3A 0E9\, 3480 rue University
SUMMARY:PhD defence of Aren Babikian – System-level testing of autonomous v
 ehicles through consistent model generation with qualitative abstractions 
 and abstract coverage
URL:https://www.mcgill.ca/ece/channels/event/phd-defence-aren-babikian-syst
 em-level-testing-autonomous-vehicles-through-consistent-model-358492
END:VEVENT
END:VCALENDAR