Published: 10th April 2018|
|Invited Presentation: Simulating and Verifying Cyber-Physical Systems: Current Challenges and Novel Research Directions Khalil Ghorbal||1|
|Automatic Generation of Communication Requirements for Enforcing Multi-Agent Safety Eric S. Kim, Murat Arcak, Sanjit A. Seshia, BaekGyu Kim and Shinichi Shiraishi||3|
|Introducing Liveness into Multi-lane Spatial Logic lane change controllers using UPPAAL Maike Schwammberger||17|
|The CAT Vehicle Testbed: A Simulator with Hardware in the Loop for Autonomous Vehicle Applications Rahul Kumar Bhadani, Jonathan Sprinkle and Matthew Bunting||32|
|Robust Safety for Autonomous Vehicles through Reconfigurable Networking Khalid Halba, Charif Mahmoudi and Edward Griffor||48|
|Identification of Risk Significant Automotive Scenarios Under Hardware Failures Mohammad Hejase, Arda Kurt, Tunc Aldemir and Umit Ozguner||59|
|On the Application of ISO 26262 in Control Design for Automated Vehicles Georg Schildbach||74|
These are the proceedings of the Second International Workshop on Safe Control of Autonomous Vehicles (SCAV 2018), which took place in Porto, Portugal on April 10th, 2018 and was part of the Cyber-Physical Systems Week 2018. Autonomous vehicles (AV) of any kind (e.g. road, maritime, aerial, unmanned) and in any configuration (e.g. individual, connected, cooperative, traffic) will provide novel services having to fulfill strong safety requirements. For the design and implementation of controllers of AVs and for control schemes of AV collectives, we identified three main objectives:
The aim of this workshop was to discuss and consolidate models, algorithms, and verification approaches for safety and resilience of the whole control loop of autonomous machines and machine collectives. The task of this workshop was to identify open research problems, discuss recent achievements, bring together researchers in, e.g. control theory, adaptive systems, machine self-organization and autonomy, mobile intelligent robotics, transportation, traffic control, machine learning, software verification, and dependability and security engineering.
The Programme Committee consisted of
We solicited both full-length papers and short papers for submission. From seven submitted full papers, five were accepted, while the only submitted short paper was accepted. Each submission was reviewed by at least three reviewers.
In addition to presentations of the contributed papers, the participants were given ample opportunity to discuss their different approaches within dedicated time slots. Furthermore, the workshop featured an invited talk by Khalil Ghorbal, entitled Simulating and Verifying Cyber-Physical Systems -- Current Challenges and Novel Research Directions.
We thank all the authors who submitted papers for considerations. Thanks go also to our invited speaker, Khalil Ghorbal. We are also very grateful to the members of the Programme Committee and additional external experts for their careful reviews.
Modeling real-life applications require the ability to combine continuous and discrete behaviors at once. The behavior of physical components, governed by electrical and kinetic laws, are naturally represented as continuous solutions of differential equations. It contrasts with several inherently discrete phenomena such as controllers' commands, modes switching, or mechanical impacts. For instance, modeling a mid-air collision avoidance system that advises the pilot to go upward or downward to avoid a detected nearby airplane requires mixing the continuous motion of the aircraft with the discrete decisions suggested by the system to resolve the conflict. The need to model, simulate and automatically analyze such systems in a compositional fashion has never been as important as nowadays, and it is perhaps one of the major challenges of the twenty first century. After a brief overview of the current state-of-the-art in modeling paradigms, simulation and verification of hybrid systems, the presentation will highlight two specific challenges, we believe, of great importance. The first is the automated generation and verification of invariant properties, that are those quantities that do not vary with respect to time. To some extent, this first challenge can be thought of as a generalization of the perhaps much familiar notion of fixed-points in the standard settings of imperative programs. The second challenge concerns the operational semantics of component-based modeling languages. Those languages are well suited to model multi-physics systems as they combine algebraic constraints, resulting from the laws of physics, in interaction with the non-smooth behavior like impact laws. Yet, the correct simulation of such hybrid models poses several subtle problems that require a careful attention. Several concrete examples will be used throughout the talk to highlight the difficulties and appreciate the several research directions we are currently exploring.