SymAware
Symbolic Logic Framework for Situational Awareness in Mixed Autonomy (SymAware)
Risk and uncertainty quantification
Temporal logic specifications
Interaction modeling of multi-agent systems
SymAware addresses the fundamental need for a new conceptual framework for awareness in multi-agent systems (MASs) that is compatible with the internal models and specifications of robotic agents and that enables the safe simultaneous operation of collaborating autonomous agents and humans. The goal of SymAware is to provide a comprehensive framework for situational awareness to support sustainable autonomy via agents that actively perceive risks and collaborate with other robots and humans to improve their awareness and understanding while fulfilling complex and dynamically changing tasks. The SymAware framework will use compositional logic, symbolic computations, formal reasoning, and uncertainty quantification to characterize and support situational awareness of MAS in its various dimensions, sustaining awareness by learning in social contexts, quantifying risks based on limited knowledge, and formulating risk-aware negotiation of task distributions. These objectives will be achieved in SymAware through (a) logical characterization of awareness using symbolic methods, (b) quantifying the symbolic reasoning for awareness with spatial and temporal ingredients for decision-making, (c) risk awareness via quantified knowledge, (d) quantifying and communicating knowledge awareness, (e) demonstrating awareness engineering in aviation and automotive use cases, and (f) identifying requirements for ethical and trustworthy awareness in human-agent interaction. The objectives of SymAware address the "Awareness Inside" Challenge of EIC by extending and formalizing human-based models of situational awareness and by providing a novel conceptual situational awareness framework for MASs that encompasses logical characterization and integrative formal reasoning of interdependent awareness dimensions including knowledge, spatiotemporal, risk and social dimensions. This will support transitioning to the safe mixed operation of autonomous agents and humans.
Period: 2022.10 - present
Role: Project leader as postdoc researcher
Principal Investigator: Dr. Sofie Haesaert
Affiliation: Eindhoven University of Technology (TU/e)
Funding Source: European Union Committee (HORIZON EIC Grants 101070802)
Website: https://www.symaware.eu/
Consortium: Max-Planck-Gesellschaft zur Forderung der Wissenschaften EV (MPI-SWS), Technische universiteit Eindhoven (TU/e), Kungliga Tekniska Hoegskolan (KTH), Uppsala Universitet (UU), Stichting Koninklijk Nederlands Lucht- en Ruimtevaartlaboratorium (NLR), Siemens Industry Software Netherlands BV.
The main objectives of SymAware include:
Logical characterization of awareness using symbolic methods.
Quantifying the symbolic reasoning for awareness with spatial and temporal ingredients for decision-making.
Risk awareness via quantified knowledge.
Quantifying and communicating knowledge awareness.
Demonstrating awareness of engineering in aviation and automotive use cases.
Identifying requirements for ethical and trustworthy awareness in human-agent interaction.
Hardware-based demonstrator
Simulation environment based on Prescan
Specification decomposition and modularized synthesis
Publications
S. Qi, Z. Zhang*, S. Haesaert, and Z. Sun, "Automated Formation Control Synthesis from Temporal Logic Specifications", 62nd IEEE Conference on Decision and Control (CDC 2023), Singapore, 13-15 Dec 2023.
Z. Zhang* and S. Haesaert, "Modularized Control Synthesis for Complex Signal Temporal Logic Specifications", 62nd IEEE Conference on Decision and Control (CDC 2023), Singapore, 13-15 Dec 2023.
L. C. Wu, Z. Zhang*, S. Haesaert, Z. Ma, and Z. Sun, "Risk-Aware Reward Shaping of Reinforcement Learning Agents for Autonomous Driving", 49th Annual Conference of the IEEE Industrial Electronics Society (IECON 2023), Singapore, 16-19 Oct 2023.
N. Dang, T. Shi, Z. Zhang, W. Jin, M. Leibold, and M.Buss, "Identifying Reaction-Aware Driving Styles of Stochastic Model Predictive Controlled Vehicles by Inverse Reinforcement Learning", 26th IEEE International Conference on Intelligent Transportation Systems (ITSC 2023), Bilbao, Spain, 24-28 Sept 2023.
Z. Zhang*, Z. Sun, and S. Haesaert, "Risk-Aware Task Assignment with Formal Specifications for Heterogeneous Multi-Agent Systems", 42nd Benelux Meeting on Systems and Control, Elspeet, The Netherlands, 21-23 Mar 2023.