Liability, Ethics, and Culture-Aware Behavior Specification using Rulebooks

@article{Censi2019LiabilityEA,
  title={Liability, Ethics, and Culture-Aware Behavior Specification using Rulebooks},
  author={Andrea Censi and Konstantin Slutsky and Tichakorn Wongpiromsarn and Dmitry S. Yershov and Scott Pendleton and James Guo Ming Fu and Emilio Frazzoli},
  journal={2019 International Conference on Robotics and Automation (ICRA)},
  year={2019},
  pages={8536-8542}
}
The behavior of self-driving cars must be compatible with an enormous set of conflicting and ambiguous objectives, from law, from ethics, from the local culture, and so on. This paper describes a new way to conveniently define the desired behavior for autonomous agents, which we use on the self-driving cars developed at nuTonomy, an Aptiv company. We define a “rulebook” as a pre-ordered set of “rules”, each akin to a violation metric on the possible outcomes (“realizations”). The rules are… Expand
From driverless dilemmas to more practical commonsense tests for automated vehicles
TLDR
This work examines, at a high level, how to test the common sense of an AV, and explains how to substantially change the premises and features of these dilemmas in order to lay the foundations for a more practical and relevant framework that tests driving common sense as an integral part of road rules testing. Expand
Safety of the Intended Driving Behavior Using Rulebooks
TLDR
The authors provide a link between the Rulebooks framework, presented by [2], and the SOTIF process and establish that Rulebooks provide a functional description of the path planning task in an AV and discuss the potential usage of the method for verification and validation. Expand
Formalization of Interstate Traffic Rules in Temporal Logic
TLDR
This work formalizes traffic rules for interstates based on the German Road Traffic Regulation, the Vienna Convention on Road Traffic, and legal decisions from courts to make it possible to automatically and unambiguously check whether traffic rules are being met by autonomous vehicles. Expand
Rules of the Road: Towards Safety and Liveness Guarantees for Autonomous Vehicles
TLDR
This paper presents a framework for the distributed control of autonomous vehicles that is safe and guarantees progress for all agents, and introduces a new game paradigm which is term the quasi-simultaneous discrete-time game. Expand
Deliberately prejudiced self-driving vehicles elicit the most outrage
TLDR
It is found that people are more outraged by AVs that kill discriminately than indiscriminately, and there may be more reason to depict and program AVs as egalitarian, insofar as the public is concerned. Expand
Rule-based optimal control for autonomous driving
TLDR
This work develops optimal control strategies for Autonomous Vehicles that are required to meet complex specifications imposed by traffic laws and cultural expectations of reasonable driving behavior, and proposes a recursive framework, in which the satisfaction of the rules in the priority structure are iteratively relaxed based on their priorities. Expand
Rule-based Evaluation and Optimal Control for Autonomous Driving
TLDR
This work develops optimal control strategies for autonomous vehicles (AVs) that are required to meet complex specifications imposed as rules of the road (ROTR) and locally specific cultural expectations of reasonable driving behavior by proposing a recursive framework, in which the satisfaction of the rules in the priority structure are iteratively relaxed in reverse order of priority. Expand
Automatic Testing and Falsification with Dynamically Constrained Reinforcement Learning
TLDR
It is proved that the agent-centric approach is able to find all dangerous behaviors that can be found by traditional falsification techniques while producing modular and reusable agents. Expand
The Reasonable Crowd: Towards evidence-based and interpretable models of driving behavior
TLDR
A dataset of 92 traffic scenarios, with multiple ways of traversing each scenario, found that the rulebook provides high interpretability without excessively sacrificing performance, and explored the interpretability vs performance trade-off by also training non-interpretable models such as a random forest. Expand
Towards Assume-Guarantee Profiles for Autonomous Vehicles
TLDR
This work proposes a systematic procedure for generating a set of supervisory specifications for self-driving cars that are associated with a distributed assume-guarantee structure and characterizable by the notion of consistency and completeness. Expand
...
1
2
3
4
...

References

SHOWING 1-10 OF 32 REFERENCES
Incorporating Ethical Considerations Into Automated Vehicle Control
TLDR
A set of ethical frameworks are used to map design decisions for a model predictive control problem to philosophical principles and the strong link between ethical principles and actual vehicle behavior developed through this approach is demonstrated. Expand
Governing lethal behavior: Embedding ethics in a hybrid deliberative/reactive robot architecture part I: Motivation and philosophy
  • R. Arkin
  • Computer Science
  • 2008 3rd ACM/IEEE International Conference on Human-Robot Interaction (HRI)
  • 2008
This paper provides the motivation and philosophy underlying the design of an ethical control and reasoning system potentially suitable for constraining lethal actions in an autonomous roboticExpand
Cake, Death, and Trolleys: Dilemmas as benchmarks of ethical decision-making
TLDR
It is argued that (some) ethical dilemmas can be used as benchmarks for estimating the ethical performance of an autonomous system and advocated that an open source repository of such dilemma should be maintained. Expand
On a Formal Model of Safe and Scalable Self-driving Cars
TLDR
A white-box, interpretable, mathematical model for safety assurance, which the authors call-Sensitive Safety (RSS), and a design of a system that adheres to the safety assurance requirements and is scalable to millions of cars. Expand
Behavior-Based Systems
TLDR
This chapter is to explain behavior-based systems and their use in autonomous control problems and applications, and provides an overview of various robotics problems and application domains that have successfully been addressed or are currently being studied with behavior- based control. Expand
Minimum-violation LTL planning with conflicting specifications
TLDR
This work proposes an algorithm for finding an optimal control strategy that maximizes the sum of rewards earned if this control strategy is applied, and demonstrates the proposed algorithm on an illustrative case study. Expand
Negotiating the probabilistic satisfaction of temporal logic motion specifications
  • Igor Cizelj, C. Belta
  • Computer Science, Mathematics
  • 2013 IEEE/RSJ International Conference on Intelligent Robots and Systems
  • 2013
We propose a human-supervised control synthesis method for a stochastic Dubins vehicle such that the probability of satisfying a specification given as a formula in a fragment of ProbabilisticExpand
Least-violating control strategy synthesis with safety rules
TLDR
This paper proposes an algorithm that synthesizes a motion which violates only lowest priority rules for the shortest amount of time, and motivates this problem with an autonomous car navigating in an urban environment while abiding by the rules of the road. Expand
Incremental sampling-based algorithm for minimum-violation motion planning
TLDR
It is proved that the proposed algorithm guarantees asymptotic optimality, i.e., almost-sure convergence to optimal solutions, on an autonomous urban mobility-on-demand system. Expand
Synthesis of Reactive Switching Protocols From Temporal Logic Specifications
We propose formal means for synthesizing switching protocols that determine the sequence in which the modes of a switched system are activated to satisfy certain high-level specifications in linearExpand
...
1
2
3
4
...