Filters








9,946 Hits in 7.2 sec

A Review of Safe Reinforcement Learning: Methods, Theory and Applications [article]

Shangding Gu, Long Yang, Yali Du, Guang Chen, Florian Walter, Jun Wang, Yaodong Yang, Alois Knoll
2022 arXiv   pre-print
Reinforcement learning has achieved tremendous success in many complex decision making tasks.  ...  When it comes to deploying RL in the real world, safety concerns are usually raised, leading to a growing demand for safe reinforcement learning algorithms, such as in autonomous driving and robotics scenarios  ...  We would like to thank Hanna Krasowski for her very useful suggestions.  ... 
arXiv:2205.10330v2 fatcat:zjfdy6t6i5f7hnb6fh5ahivz5e

Safe Policy Synthesis in Multi-Agent POMDPs via Discrete-Time Barrier Functions

Mohamadreza Ahmadi, Andrew Singletary, Joel W. Burdick, Aaron D. Ames
2019 2019 IEEE 58th Conference on Decision and Control (CDC)  
A multi-agent partially observable Markov decision process (MPOMDP) is a modeling paradigm used for high-level planning of heterogeneous autonomous agents subject to uncertainty and partial observation  ...  In this paper, we use barrier functions to design policies for MPOMDPs that ensure safety. Notably, our method does not rely on discretizations of the belief space, or finite memory.  ...  Multi-agent partially observable Markov decision processes [24] , [8] provide a sequential decision-making formalism for high-level planning of multiple autonomous agents under partial observation and  ... 
doi:10.1109/cdc40024.2019.9030241 dblp:conf/cdc/AhmadiSBA19 fatcat:j2bptoefdjaq5mxidjj55mtxu4

Safe Policy Synthesis in Multi-Agent POMDPs via Discrete-Time Barrier Functions [article]

Mohamadreza Ahmadi, Andrew Singletary, Joel W. Burdick, Aaron D. Ames
2019 arXiv   pre-print
A multi-agent partially observable Markov decision process (MPOMDP) is a modeling paradigm used for high-level planning of heterogeneous autonomous agents subject to uncertainty and partial observation  ...  In this paper, we use barrier functions to design policies for MPOMDPs that ensure safety. Notably, our method does not rely on discretization of the belief space, or finite memory.  ...  Multi-agent partially observable Markov decision processes [26] , [10] provide a sequential decision-making formalism for high-level planning of multiple autonomous agents under partial observation  ... 
arXiv:1903.07823v2 fatcat:yf5evbbzfvhrfkmrmu6lmfcrbe

Meeting the challenges of decentralised embedded applications using multi-agent systems

Jean Paul Jamont, Michel Occello
2015 International Journal of Agent-Oriented Software Engineering  
They require a decentralized embedded intelligence generating challenges for embedded systems. A multi-agent approach is well suited to model and design decentralized embedded applications.  ...  Reference Jean-Paul Jamont and Michel Occello (20xx) Meeting the challenges of decentralized embedded applications using multi-agent systems, Int.  ...  Works are currently done to apply autonomic systems to embedded systems as it is done for multi-agent systems (Chun et al., 2010) .  ... 
doi:10.1504/ijaose.2015.078435 fatcat:2av2ehrmijdijkm5vdqa2o5zv4

Using Agent-Based Modelling Approaches to Support the Development of Safety Policy for Systems of Systems [chapter]

Martin Hall-May, Tim Kelly
2006 Lecture Notes in Computer Science  
A safety policy defines the set of rules that governs the safe interaction of agents operating together as part of a system of systems (SoS).  ...  Methods for multi-agent system design can aid in this understanding. Such approaches mention organisational rules. However, there is little discussion about how they are derived.  ...  Introduction Making Systems of Systems Safe A system of systems (SoS) is a large-scale network of autonomous, heterogeneous, and often mobile entities that are individually purposeful, and yet are expected  ... 
doi:10.1007/11875567_25 fatcat:lvh722s4o5fqxlhdh3cnsyk72a

Dynamic security reconfiguration for the semantic web

Juan Jim Tan, Stefan Poslad
2004 Engineering applications of artificial intelligence  
a security reconfiguration, and to decide if a suitable level of security interoperability between heterogeneous systems is achievable.  ...  Security requirements are defined using security profiles that describe the interlinking of security policies to instances of services.  ...  Decision making policies deduce decisions from the general security policies (Section 4.1.1).  ... 
doi:10.1016/j.engappai.2004.08.036 fatcat:l57i4legf5adtoh2njm6jtoqvu

Dynamic security reconfiguration for the semantic web

J TAN, S POSLAD
2004 Engineering applications of artificial intelligence  
a security reconfiguration, and to decide if a suitable level of security interoperability between heterogeneous systems is achievable.  ...  Security requirements are defined using security profiles that describe the interlinking of security policies to instances of services.  ...  Decision making policies deduce decisions from the general security policies (Section 4.1.1).  ... 
doi:10.1016/s0952-1976(04)00118-6 fatcat:fnd2pzwctncxrb5zhug5yf4jri

Recent Advances in Deep Reinforcement Learning Applications for Solving Partially Observable Markov Decision Processes (POMDP) Problems Part 2—Applications in Transportation, Industries, Communications and Networking and More Topics

Xuanchen Xiang, Simon Foo, Huanyu Zang
2021 Machine Learning and Knowledge Extraction  
Reinforcement Learning (RL) is an approach to simulate the human's natural learning process, whose key is to let the agent learn by interacting with the stochastic environment.  ...  The two-part series of papers provides a survey on recent advances in Deep Reinforcement Learning (DRL) for solving partially observable Markov decision processes (POMDP) problems.  ...  The adversarial multi-agent policy improves system efficiency even under stochastic. Based on this, Jang et al.  ... 
doi:10.3390/make3040043 doaj:45bf00de595c44d186fa3d200589c1c5 fatcat:qx4srh7qabgjvd5l6lj6nulhxa

Practical Challenges in Explicit Ethical Machine Reasoning [article]

Louise Dennis, Michael Fisher
2018 arXiv   pre-print
We examine implemented systems for ethical machine reasoning with a view to identifying the practical challenges (as opposed to philosophical challenges) posed by the area.  ...  We identify a need for complex ethical machine reasoning not only to be multi-objective, proactive, and scrutable but that it must draw on heterogeneous evidential reasoning.  ...  We note that this makes ethical reasoning inherently multi-objective which is a practical challenge to many techniques for controlling decision-making in machines.  ... 
arXiv:1801.01422v1 fatcat:63nhcoxowzb3rdb4hkaaktgxom

The social behavior of autonomous vehicles

Lars Müller, Malte Risto, Colleen Emmenegger
2016 Proceedings of the 2016 ACM International Joint Conference on Pervasive and Ubiquitous Computing Adjunct - UbiComp '16  
We model interactions between human and autonomous agents with game theory and the principle of best response.  ...  We integrate tools from social psychology into autonomous-vehicle decision making to quantify and predict the social behavior of other drivers and to behave in a socially compliant way.  ...  We thank them for this support.  ... 
doi:10.1145/2968219.2968561 dblp:conf/huc/MullerRE16 fatcat:nryythr6p5bcjavhopwvpo6eji

Analysis, Optimization, Control, and Learning of Cyber-Physical Systems [article]

Andreas Malikopoulos
2022 arXiv   pre-print
to developing rigorous mathematical models and decentralized control algorithms for making engineering complex systems able to realize how to improve their performance over time while interacting with  ...  The overarching goal of the Information and Decision Science (IDS) Lab is to enhance understanding of complex systems and establish a holistic, multifaceted approach using scalable data and informatics  ...  Multi-Agent and Swarm System There are two application areas that we have identified to study and control emergent behavior in multi-agent systems.  ... 
arXiv:2109.09055v2 fatcat:5obayt63tzgihmn3k5uucrsb2i

Reactive and Safe Road User Simulations using Neural Barrier Certificates [article]

Yue Meng, Zengyi Qin, Chuchu Fan
2021 arXiv   pre-print
In this work, we proposed a reactive agent model which can ensure safety without comprising the original purposes, by learning only high-level decisions from expert data and a low-level decentralized controller  ...  Reactive and safe agent modelings are important for nowadays traffic simulator designs and safe planning applications.  ...  The views, opinions, and/or findings expressed are those of the authors and should not be interpreted as representing the official views or policies of any NASA entity, DSTA Singapore, or the Singapore  ... 
arXiv:2109.06689v1 fatcat:5pwdfvc2jzb4pjay474cimn3m4

In-Time Explainability in Multi-Agent Systems: Challenges, Opportunities, and Roadmap [chapter]

Francesco Alzetta, Paolo Giorgini, Amro Najjar, Michael I. Schumacher, Davide Calvaresi
2020 Lecture Notes in Computer Science  
On the one hand, systems dealing with strict-timing constraints in safety-critical applications mainly focused on predictability, leaving little room for complex planning and decisionmaking processes.  ...  In particular, it proposes to embrace the Real-Time Beliefs Desires Intentions (RT-BDI) framework as an enabler of eXplainable Multi-Agent Systems (XMAS) in time-critical XAI.  ...  BDI agents and XMAS can both make autonomous decisions dynamically.  ... 
doi:10.1007/978-3-030-51924-7_3 fatcat:cqvb3aejjbdphea4h6j7rmyaca

Learn-to-Race Challenge 2022: Benchmarking Safe Learning and Cross-domain Generalisation in Autonomous Racing [article]

Jonathan Francis, Bingqing Chen, Siddha Ganju, Sidharth Kathpal, Jyotish Poonganam, Ayush Shivani, Vrushank Vyas, Sahika Genc, Ivan Zhukov, Max Kumskoy, Anirudh Koul, Jean Oh (+1 others)
2022 arXiv   pre-print
Analogous to racing being used to test cutting-edge vehicles, we envision autonomous racing to serve as a particularly challenging proving ground for autonomous agents as: (i) they need to make sub-second  ...  In the first stage of the challenge, we evaluate an autonomous agent's ability to drive as fast as possible, while adhering to safety constraints.  ...  We also thank the Challenge participants themselves for their engagement and feedback-in particular, the lachlan mares and matthew howe participants for contributing their approach descriptions.  ... 
arXiv:2205.02953v2 fatcat:5dnbydujejckdgmbbw5kyer46u

Social behavior for autonomous vehicles

Wilko Schwarting, Alyssa Pierson, Javier Alonso-Mora, Sertac Karaman, Daniela Rus
2019 Proceedings of the National Academy of Sciences of the United States of America  
We integrate tools from social psychology into autonomous-vehicle decision making to quantify and predict the social behavior of other drivers and to behave in a socially compliant way.  ...  This approach allows autonomous vehicles to observe human drivers, estimate their SVOs, and generate an autonomous control policy in real time.  ...  We thank them for this support.  ... 
doi:10.1073/pnas.1820676116 pmid:31757853 pmcid:PMC6911195 fatcat:mlz7ciesr5gkpo4jf4vsjyk5cy
« Previous Showing results 1 — 15 out of 9,946 results