|
Published: 30th September 2023 DOI: 10.4204/EPTCS.391 ISSN: 2075-2180 |
This volume encompasses the proceedings of the Third Workshop on Agents and Robots for reliable Engineered Autonomy (AREA 2023), co-located with the 26th European Conference on Artificial Intelligence (ECAI 2023).
The realm of autonomous agents, extensively studied for decades, has been a focal point from both design and implementation standpoints. Nevertheless, the practical utilization of agents in real-world scenarios has predominantly been within software-centric applications, with limited adoption in situations necessitating physical interactions. Concurrently, the utility of robots has transcended narrow industrial contexts and has expanded across various domains. These domains span from robotic assistants to search and rescue operations, wherein the operational context is dynamic and not fully specified. This context often involves intricate interactions between multiple robots and humans.
These circumstances pose notable challenges to conventional software engineering methods. Enhanced autonomy stands as a pivotal avenue for enabling effective functioning of robotic applications in such settings. Autonomous agents and multi-agent systems emerge as promising methodologies for their development. As the levels of autonomy and interaction escalate, ensuring reliable behavior becomes increasingly intricate, not only in robotic applications but also in conventional autonomous agent scenarios. Hence, there exists a demand for research into novel verification and validation approaches that can be seamlessly integrated into the developmental life cycle of these systems.
The primary objective of this workshop is to facilitate collaboration between researchers in the fields of autonomous agents and robotics. By amalgamating insights from these domains, innovative solutions could potentially emerge to tackle intricate challenges associated with the verification and validation of autonomous robotic systems.
In this third iteration of the workshop, a total of 12 submissions were received, of which 7 full papers and 5 short papers were accepted. We extend our gratitude to all authors who contributed their valuable work to the workshop.
Finally, we would like to thank our invited speakers, Viviana Mascardi and Bruno Lacerda. The title and abstract of their presentations can be found below.
We also express our appreciation to the 25 program committee members (the complete list is available below) for their valuable feedback, which contributed to the refinement of the papers. Our thanks also extend to the EPTCS staff for their support in compiling these proceedings.
For more information about the workshop, please check our website: https://areaworkshop.github.io/AREA2023/
The AREA 2023 organisers,
Angelo Ferrando and Rafael C. Cardoso
What do chatbots and the metaverse have to do with reliability, cognitive agents, and robotic applications? In this talk I will explore the intriguing connections among them, presenting a small scale prototype of (reliable) "talking agents in the virtual world".
In this presentation, I will argue that the synergy between three factors is critical for creating reliable mission planning algorithms for autonomous robots operating in uncontrolled environments. These factors are (i) utilising historical data gathered online to enhance decision-making under uncertainty models, (ii) implementing principled planning techniques that explicitly reason about the epistemic uncertainty inherent to these data-driven models, and (iii) incorporating rich specifications that go beyond typical expected reward maximisation problems. Developing frameworks that unify these three factors is an open problem in the field of robotics, which is heavily dependent on the specific application domain. I will offer an overview of various works undertaken at the GOALS lab at the Oxford Robotics Institute that consider these three factors in distinct ways. Moreover, I will describe how these works provide a foundation for creating integrated approaches that enable long-term deployment of robots capable of acquiring parametrised environmental models from historical data, plan considering the epistemic uncertainty of such models, and effectively adapt their behaviour to in-mission observations, continuously refining their estimate over the epistemic uncertainty online.