Learning product set models of fault triggers in high-dimensional software interfaces

Vernaza, Paul; Guttendorf, David; Wagner, Michael; Koopman, Philip

doi:10.1109/iros.2015.7353866

Cited by 6 publications

(2 citation statements)

References 15 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…The potential benefits of using FI into design phases of autonomous systems range from providing early opportunities for integration of inductive technologies-e.g., machine learning algorithms that use training sets to derive models of camera lens-to reducing costs and risks associated to autonomy functions. Such techniques have already been used successfully to find and characterize defects on autonomous vehicles [36].…”

Section: Towards Robustness Validation Approach For Autonomymentioning

confidence: 99%

Sharpening the Scythe of Technological Change: Socio-Technical Challenges of Autonomous and Adaptive Cyber-Physical Systems

Cancila

Gerstenmayer²,

Espinoza

2018

Designs

View full text Add to dashboard Cite

Autonomous and Adaptative Cyber-Physical Systems (ACPS) represent a new knowledge frontier of converging “nano-bio-info-cogno” technologies and applications. ACPS have the ability to integrate new `mutagenic’ technologies, i.e., technologies able to cause mutations in the society. Emerging approaches, such as artificial intelligence techniques and deep learning, enable exponential speedups for supporting increasingly higher levels of autonomy and self-adaptation. In spite of this disruptive landscape, however, deployment and broader adoption of ACPS in safety-critical scenarios remains challenging. In this paper, we address some challenges that are stretching the limits of ACPS safety engineering, including tightly related aspects such as ethics and resilience. We argue that a paradigm change is needed that includes the entire socio-technical aspects, including trustworthiness, responsibility, liability, as well as the ACPS ability to learn from past events, anticipate long-term threads and recover from unexpected behaviors.

show abstract

Section: Towards Robustness Validation Approach For Autonomymentioning

confidence: 99%

Sharpening the Scythe of Technological Change: Socio-Technical Challenges of Autonomous and Adaptive Cyber-Physical Systems

Cancila

Gerstenmayer²,

Espinoza

2018

Designs

View full text Add to dashboard Cite

show abstract

“…Such techniques have already been used successfully to find and characterize defects on autonomous vehicles. [35] A promising approach to helping validate autonomy features is to perform fault injection at the level of abstraction of the component, as part of a strategy of attempting to falsify claims of safety. [36] This involves not only simulating objects for primary sensor inputs, but also inserting exceptional conditions to test the robustness of the system (e.g., inserting invalid data into maps).…”

Section: Fault Injectionmentioning

confidence: 99%

Challenges in Autonomous Vehicle Testing and Validation

Koopman

Wagner

2016

SAE Int. J. Trans. Safety

Self Cite

485

297

View full text Add to dashboard Cite

Software testing is all too often simply a bug hunt rather than a wellconsidered exercise in ensuring quality. A more methodical approach than a simple cycle of system-level test-fail-patch-test will be required to deploy safe autonomous vehicles at scale. The ISO 26262 development V process sets up a framework that ties each type of testing to a corresponding design or requirement document, but presents challenges when adapted to deal with the sorts of novel testing problems that face autonomous vehicles. This paper identifies five major challenge areas in testing according to the V model for autonomous vehicles: driver out of the loop, complex requirements, non-deterministic algorithms, inductive learning algorithms, and failoperational systems. General solution approaches that seem promising across these different challenge areas include: phased deployment using successively relaxed operational scenarios, use of a monitor/actuator pair architecture to separate the most complex autonomy functions from simpler safety functions, and fault injection as a way to perform more efficient edge case testing. While significant challenges remain in safety-certifying the type of algorithms that provide high-level autonomy themselves, it seems within reach to instead architect the system and its accompanying design process to be able to employ existing software safety approaches.

show abstract