WOSRA

Towards a Robotic Society

Day by day, new machines and systems are being developed to help and assist humans in a myriad of activities. Future robotic systems need to work in teams and communicate with humans and other robots to share information and coordinate activities. In particular, there is an increasing demand from government agencies and the private sector alike to use Unmanned Aerial Vehicles (UAVs), Unmanned Ground Vehicles (UGVs), Unmanned Surface Vehicles (USVs), and Autonomous Underwater Vehicles (AUVs) for tasks including homeland security, reconnaissance, search and rescue, surveillance, data collection, and urban planning. As these systems will interact with humans in several scenarios, it is urgent to discuss the ethical aspects in their development related to, for instance, accountability, privacy and data protection, alignment to human values in both design and function. Ontology-based models for this domain would enable, for example, a clear communication among different stakeholders, the formulations of laws, the building of AI-based and Robotics systems with full alignment with what stakeholders expect, in terms of benefits and increase of human well-being.

Objectives

Context

Robotics is becoming a mainstream domain with a wide range of applications with middle and long-term impact on everyone lives. Current systems rely more and more on robot-robot communication and robot-human interaction. In terms of communication, a vocabulary with clear and concise definitions is a sine qua non component to enable information exchange among any group of agents, which can be human or non-human actors. This need for a well-defined knowledge representation is becoming evident if one considers the growing complexity of behaviors that robots are expected to perform as well as the rise of multi-robot and human-robot collaboration. The existence of a standard knowledge representation would:

• define precisely concepts and relations in the robot's knowledge representation that could include, but is not limited to, robot hardware and software, environment, cause and effects of performing actions, relationship among other robots and people;
• ensure common understanding among members of the community;
• facilitate efficient data integration and transfer of information among robotic systems.

Objectives

This workshop has the purpose to increase interest in standardization for the Robotics and Automation (R&A) domain, as well as the ethical challenges involved. Our previous work in this context resulted in the very first RAS standard, named IEEE 1872-2015 - IEEE Standard Ontologies for Robotics and Automation. It defined a core ontology to specify the main concepts, relations, and axioms of robotics and automation and has become a reference for knowledge representation and reasoning in robots, as well as a formal reference vocabulary for communicating knowledge about R&A between robots and humans.

As extensions to the IEEE 1872-2015 standard, new standards are under development. The Standards Development Working Group on Robot Task Representation (RTR) is currently working is to provide a robot task ontology for knowledge representation and reasoning in robotics and automation. Likewise, the Standards Development Working Group on Standard for Autonomous Robotics (AuR) Ontology works towards a unified way of representing autonomous robotics architectures across different R&A domains, to provide a robot with autonomy. Aligned with these efforts, the IEEE P7007 Ontological Standard for Ethically Driven Robotics and Automation Systems WG activities address the main issues concerning the development of Ethical AI and Robotics systems with special focus on ontology-based approach to define the foundation of ethics concepts and terminology for AI, R&A systems in accordance with Ethics and Moral theories.

This workshop aims to discuss this emergent domain through invited talks, short presentations, and application papers from different areas with two main objectives:

1. highlight the recent developments in ontology-based approaches to describe the robotics and automation domain in terms of general knowledge, applications and ethical issues;
2. share and compare different viewpoints from robotics practitioners, looking for a common ground to combine distinct approaches towards the development of standards as part of the IEEE Standard Projects.

The workshop is structured to elicit participation and feedback from a broad spectrum of participants, both from academia and industry, interested in advancing the interaction between robots and humans using a standardized and formal terminology. The proposed workshop will serve the following sectors: Public sector using autonomous robots to deliver automated services, such as police and emergency responders; Health service providers; Transformation industry; Mining and construction industry; Oil and gas industry.

Topics of interest

• Autonomous Robotics
• Ontology-based development for Robotics and Automation
• Ontology-based Standards for Robotics and Automation
• Knowledge representation and reasoning for robotics and automation using Ontologies
• Robot-robot interaction and/or Human-robot interaction using ontologies
• Building ethical AI and Robot systems
• Ethical design and development of AI and Robot systems
• Data privacy and protection within world of AI and Robots
• Trust in AI and Robots
• Fairness and Transparency in AI and Robots
• AI, Robotics and law

Intended audience

The targeted audience covers a wide range of practitioners from and outside Robotics that include (but not limited) people from Computer Science, Robotics, Philosophy, Engineering, and Law. We are specifically interested to reach participants from industrial and academic organisations, and early career researchers such as graduate students and postdoctoral researchers, interested on how ontologies and the future standard can be used to support their developments and research on autonomous robotics.

Workhsop content

Structure of the event

The event will be structured around sessions of invited speakers separated by coffee and lunch breaks. Prominent speakers and well-known experts in the field will concentrate in this workshop to better disseminate our growing community among the IROS18 attendees. One of the sessions will also include a significant time for interactive sessions and open-ended discussions.

There will be four sessions with invited speakers and experts in the topics covered by the Workshop. One session will be devoted to short presentations coming mainly from young researchers and PhD students, and also industry actors in the field of robotics and artificial intelligence with five-minute slots. Participants will be able to lively discuss the current trends and future challenges based on the case studies presented by each invited speaker in the proposed sessions.

Workshop results will be disseminated afterwards. Abstracts will be made available online, on the Workshop Website. Presentations and discussions will be videotaped and uploaded to the workshop website, under permission. Consensus documents and the transcript of the discussions will become supporting materials of the ontological standards and will be made available at the website. Active participants not already involved in the Working Groups will be invited to join.

Invited Speakers

Talk 1: Stefano Borgo, Laboratory for Applied Ontology, ISTC/ CNR, Italy.
Talk 2: Michael Beetz, Institute for Artificial Intelligence (IAI), University of Bremen, Germany.
Talk 3: Ricardo Sanz, Autonomous Systems Laboratory, Universidad Politécnica de Madrid, Spain.
Talk 4: Howard Li, Department of Electrical and Computer Engineering, University of New Brunswick, Fredericton, New Brunswick.
Talk 5: Joanna Isabelle Olszewska, University of West Scotland, UK.
Talk 6: Paulo Gonçalves, IDMEC - Center of Intelligent Systems, University of Lisbon, Portugal.
Talk 7: Edison Pignaton de Freitas, Federal University of Rio Grande do Sul Informatics Institute, Brazil.
Talk 8: Paulo Menezes, University of Coimbra, Portugal.
Talk 9: Edson Prestes, Informatics Institute, Federal University of Rio Grande do Sul, Brazil.
Talk 10: Hagita Nori, ATR Lab, Japan.
Talk 11: Naouel Ayari, E-Helth/Research Departement, Altran Technologies, France.
Talk 12: S. Veera Ragavan, Department of Mechatronics Engineering, School of Engineering, Monash University (Malaysia Campus)

Abstracts

KnowRob 2.0 - A 2nd Generation Knowledge Processing Framework for Cognition-enabled Robotic Agents

Prof. Michael Beetz, Director

The Institute of Artificial Intelligence (IAI), Bremen, Germany

In this talk I present KnowRob2.0, a second generation knowledge representation and reasoning framework for robotic agents. KnowRob2.0 is an extension and partial redesign of KnowRob, currently one of the most advanced knowledge processing systems for robots that has enabled them to successfully perform complex manipulation tasks such as making pizza, conducting chemical experiments, and setting tables. The knowledge base appears to be a conventional first-order time interval logic knowledge base, but it exists to a large part only virtually: many logical expressions are constructed on demand from data structures of the control program, computed through robotics algorithms including ones for motion planning and solving inverse kinematics problems, and log data stored in noSQL databases. Novel features and extensions of KnowRob2.0 substantially increase the capabilities of robotic agents of acquiring open-ended manipulation skills and competence, reasoning about how to perform manipulation actions more realistically, and acquiring commonsense knowledge.

Engineering modeling and ontological modeling: God and devil are in the detail

Stefano Borgo

Laboratory for Applied Ontology, ISTC CNR, Italy

The increasing interest to build comprehensive models in engineering and robotics pushes for the adoption of reliable methodologies and, among these, semantic and ontological approaches are the most promising. There is however a tension in attitude and principle between engineering, semantic and ontological modeling and, because of this, research focus and solutions diverge. On the one hand researchers are pushing for "translating" existing engineering models and standards into semantic languages like OWL. On the other hand, researchers are rebuilding innovative models and standards based on ontological analysis. The latter approach can lead to change how we look at engineering data and problems. There are different ways to proceed and we should be aware of the differences. We will give examples of differences between engineering and ontological models, and also discuss how the choice of the ontology can impact your work in robotics.

Development of Autonomous Robotics Ontology

Howard Li

University of New Brunswick. Autonomous Robotics (AuR) Ontology IEEE Working Group Chair

The IEEE SA Autonomous Robotics Ontology Working Group aims to extend the CORA ontology to represent more specific concepts and axioms that are commonly used in autonomous robotics. Therefore, the autonomous robotics working group performs a comprehensive study in different R&A domains (e.g. aerial, ground, surface, underwater, and space robots) to identify the basic hardware and software components necessary to provide a robot (or a group of robots) with autonomy. As a long-term goal, the working group will create a standard ontology that specifies the domain knowledge needed to build autonomous systems comprised of robots that can operate in all classes of unstructured environments. The autonomous robotics working group aims to extend the CORA ontology for the domain of autonomous robotics by:

1. Studying R&A sub-domains to identify the core design patterns specific to autonomous robotics.
2. Developing general ontological concepts and domain-specific axioms for autonomous robotics.
3. Analyze and compare existing systems in order to develop general use cases and/or case studies for design of autonomous robotics.

Sharing Knowledge in the Intelligent Robot Life-cycle

Ricardo Sanz

Autonomous Systems Laboratory / Center for Automation and Robotics. Universidad Politecnica de Madrid, Spain

In the life-cycles of intelligent robots there are two knowledge-based systems of maximal relevance: the intelligent robot system and the engineering system. The intelligent robot uses knowledge in the performance of its tasks and in its relations with other agents. The engineering system is the human-centric, tool-intensive system that produces the robot system. This system uses the knowledge to build, operate, maintain and evolve the robot system. Both systems are obviously causally related but the coupling at the knowledge level is of special interest. In this talk I will address the nature and roles of this knowledge in an MBSE life-cycle context and how ontologies are the cornerstones for all this process.

Ontologies for Vision Agents

Dr Joanna Isabelle Olszewska

University of West Scotland, UK

With the growth of AI agents in our daily life, people are increasingly asked to interact with these intelligent agents, which are most of the time equipped with camera(s) and/or systems able to automatically process visual inputs such as raw still images, live video feeds, etc. Hence, the development of ontologies for these vision agents is of prime importance, in order to allow the sharing of the information between humans and/or vision-enabled robotic systems in a consistent and interoperable way. In this work, we present the challenges underlying vision agents as well as the solutions provided by the development of ontologies such as the Spatio-Temporal Visual Ontology (STVO).

Ontologies for AAL and the integration of Artificial Social Agents

Paulo Menezes

University of Coimbra, Portugal

In this talk, we will address some recent efforts regarding the representation of knowledge associated with the Active and Assisted Living (AAL) domain and the approaches that are being proposed to integrate such knowledge into Artificial Social Agents. In particular, we will look into the recently proposed taxonomy for AAL - TAALxonomy - and what concepts are aligned with the current project for standardization regarding ontologies for autonomous systems, with particular emphasis on the Core Ontology for Robotics and Automation (CORA). We will also briefly present an application example of how these knowledge representations are helping researchers to develop more reliable artificial social agents.

Enhancing Information Sharing in Cloud-Robotics Systems by using Ontologies

Edison Pignaton de Freitas

Federal University of Rio Grande do Sul

The usage of cloud-based systems represents a no-way-back trend, which is allowing the further developments in different areas. In robotics domain it is also true, as cloud-robotics systems are emerging as a suitable solution model for cooperative robotics systems composed of groups of robots that interface and cooperate with computer systems. Exploring this topic, this talk presents efforts performed by the IEEE SA Autonomous Robotics Ontology Working Group to address the need for an ontology to this specific domain, highlighting the main challenges to be faced in the near future.

Ontologies for Healthcare Robots: development and applications

Paulo J.S. Gonçalves

IDMEC - Center of Intelligent Systems, University of Lisbon, Portugal

The talk presents an overview of the efforts (current and past) to deploy healthcare robots in healthcare scenarios, that use ontologies to represent knowledge. Several scenarios tackled by the research community uses ontologies in robotic systems to share knowledge, in a machine readable format, to solve specific tasks in healthcare sub-domains as: surgical robots, robot companions, and so on. Such surveyed domain specific ontologies will be presented in the talk. After a discussion will be performed about the upper level ontologies that are used in those cases, which often are in conflict, e.g. BFO used mainly for the biomedical domain and SUMO used in IEEE-CORA. The presentation will end by presenting the development workflow of a sub-domain ontology: Ontology for Robotic Orthopaedic Surgery (OROSU).

N-ary Ontologies-based Models for Assistive Robotics: Challenges & Future Trends

Naouel Ayari

E-Helth/Research Departement, Altran Technologies, France

The next generation of ambient assisted living services will be based on eco-systems or organizations of intelligent artificial agents embodied in companion robots and smart objects. To provide, anywhere and anytime, smart assistance services to people, these agents need to be endowed with advanced knowledge representation, reasoning and communication capabilities. To render assistive agents more autonomous and able of taking context-aware decisions based on the current context/situation of the assisted person, expressive and efficient knowledge representation and reasoning models for the open world are required. I will talk about recent work in joint reasoning about semantic representations based on n-ary ontologies and physical representations, especially how such reasoning relates to natural language understanding, and how we can bridge the gap between low-level sensing and control, and higher-level semantic representations to create more capable robots for healthcare applications.

Contact