Design of a Knowledge-Base Strategy for Capability-Aware Treatment of Uncertainties of Automated Driving Systems

Automated Driving Systems (ADS) represent a key technological advancement in the area of Cyber-physical systems (CPS) and Embedded Control Systems (ECS) with the aim of promoting traffic safety and environmental sustainability. The operation of ADS however exhibits several uncertainties that if improperly treated in development and operation would lead to safety and performance related problems. This paper presents the design of a knowledge-base (KB) strategy for a systematic treatment of such uncertainties and their system-wide implications on design-space and state-space. In the context of this approach, we use the term Knowledge-Base (KB) to refer to the model that stipulates the fundamental facts of a CPS in regard to the overall system operational states, action sequences, as well as the related costs or constraint factors. The model constitutes a formal basis for describing, communicating and inferring particular operational truths as well as the belief and knowledge representing the awareness or comprehension of such truths. For the reasoning of ADS behaviors and safety risks, each system operational state is explicitly formulated as a conjunction of environmental state and some collective states showing the ADS capabilities for perception, control and actuations. Uncertainty Models (UM) are associated as attributes to such state definitions for describing and quantifying the corresponding belief or knowledge status due to the presences of evidences about system performance and deficiencies, etc. On a broader perspective, the approach is part of our research on bridging the gaps among intelligent functions, system capability and dependability for mission-&safety-critical CPS, through a combination of development- and run-time measures.

[1]  Tatsuya Suzuki,et al.  Analyzing driver gaze behavior and consistency of decision making during automated driving , 2015, 2015 IEEE Intelligent Vehicles Symposium (IV).

[2]  Peter H. Feiler,et al.  Model-Based Engineering with AADL: An Introduction to the SAE Architecture Analysis & Design Language , 2012 .

[3]  David J. C. MacKay,et al.  Information Theory, Inference, and Learning Algorithms , 2004, IEEE Transactions on Information Theory.

[4]  John A. Michon,et al.  A critical view of driver behavior models: What do we know , 1985 .

[5]  Lars Grunske,et al.  Evaluating probabilistic models with uncertain model parameters , 2012, Software & Systems Modeling.

[6]  Frederick Hayes-Roth,et al.  Building expert systems , 1983, Advanced book program.

[7]  Sara Tucci Piergiovanni,et al.  Model-Based Analysis and Engineering of Automotive Architectures with EAST-ADL: Revisited , 2015, Int. J. Concept. Struct. Smart Appl..

[8]  Enrico Zio,et al.  Uncertainty in Risk Assessment: The Representation and Treatment of Uncertainties by Probabilistic and Non-Probabilistic Methods , 2013 .

[9]  Bran Selic,et al.  Understanding Uncertainty in Cyber-Physical Systems: A Conceptual Model , 2016, ECMFA.

[10]  Xueqin Zhang,et al.  Support Vector Machines for Anomaly Detection , 2006, 2006 6th World Congress on Intelligent Control and Automation.

[11]  James S. Albus,et al.  A Reference Model Architecture for Intelligent Hybrid Control Systems , 1996 .

[12]  DeJiu Chen,et al.  A Methodological Framework for Model-Based Self-management of Services and Components in Dependable Cyber-Physical Systems , 2017, DepCoS-RELCOMEX.

[13]  Krishna R. Pattipati,et al.  A hidden Markov model-based algorithm for fault diagnosis with partial and imperfect tests , 2000, IEEE Trans. Syst. Man Cybern. Part C.

[14]  Rolf Johansson,et al.  A Strategy for Assessing Safe Use of Sensors in Autonomous Road Vehicles , 2017, SAFECOMP.

[15]  A. Kiureghian,et al.  Aleatory or epistemic? Does it matter? , 2009 .

[16]  Fausto Giunchiglia,et al.  NUSMV: a new symbolic model checker , 2000, International Journal on Software Tools for Technology Transfer.