Recent & Upcoming Talks

07/11/17

Transparency in AI: Aiming at understanding the machines and then humans
Experiments, Morals, and Machines Workshop Lille Catholic University, Lille, France

01/18/17

Why is my Roomba trying to kill my cat? The need to build transparent machines
University of Cyprus, Nicosia, Cyprus

11/29/16

Why is my Roomba trying to kill my cat? The need to build transparent machines
Georgia Institute of Technology, Atlanta GA, USA

09/24/16

Transparency as a Consideration in Building AI
Responsible AI PhD Summer School, The Hague, Netherlands

08/30/16

Transparency as an Ethical Consideration
ECAI 16: Ethics in Design of Intelligent Agents Workshop, The Hague, Netherlands

05/05/16

Transparency in Intelligence
Computer Science PhD Conference, Bath, UK

04/04/16

Why is my robot behaving like that?
AISB 16: Symbosia in EPSRC Principle of Robotics, Sheffield, UK

Journal Articles

Designing and implementing transparency for real time inspection of autonomous robots

Andreas Theodorou, Robert R. Wortham, and Joanna J. Bryson
Connection Science: Vol.: 29 Issue: 3 2017

Expand abstractThe EPSRC’s Principles of Robotics dictates the implementation of transparency in robotic systems, however, research related to it is in its infancy The current paper introduces the reader to the need of having transparent to inspection intelligent agents. We provide a robust definition of transparency, as a mechanism to expose the decision making of the robot, by considering and expanding upon other prominent definitions found in literature. The paper concludes by addressing potentials design decisions developers need to consider when designing transparent systems.

University of Bath OPUS, Taylor & Francis Online

RobotTransparency,Trust and Utility

Robert R. Wortham and Andreas Theodorou
Connection Science: Vol.: 29 Issue: 3 2016

Expand abstractThe EPSRC’s Principles of Robotics dictates the implementation of transparency in robotic systems, however, research related to it is in its infancy. The current paper introduces the reader to the need of having transparent to inspection intelligent agents. We provide a robust definition of transparency, as a mechanism to expose the decision making of the robot, by considering and expanding upon other prominent definitions found in literature. The paper concludes by addressing potentials design decisions developers need to consider when designing transparent systems.

University of Bath OPUS, Taylor & Francis Online

Conference Publications

Improving Robot Transparency: Real-Time Visualisation of Robot AI Substantially Improves Understanding in Naive Observers

Robert R. Wortham, Andreas Theodorou, and Joanna J. Bryson
RO-MAN 2017 2017

Expand abstractDeciphering the behaviour of intelligent others is a fundamental characteristic of our own intelligence. As we interact with complex intelligent artefacts, humans inevitably construct mental models to understand and predict their behaviour. If these models are incorrect or inadequate, we run the risk of self deception or even harm. Here we demonstrate that providing even a simple, abstracted real-time visualisation of a robot’s AI can radically improve the transparency of machine cognition. Findings from both an online experiment using a video recording of a robot, and from direct observation of a robot show substantial improvements in observers’ understanding of the robot’s behaviour. Unexpectedly, this improved understanding was correlated in one condition with an increased perception that the robot was ‘thinking’, but in no conditions was the robot’s assessed intelligence impacted. In addition to our results, we describe our approach, tools used, implications, and potential future research directions.

University of Bath OPUS

Robot Transparency: Improving Understanding of Intelligent Behaviour for Designers and Users

Robert R. Wortham, Andreas Theodorou, and Joanna J. Bryson
TAROS 2017 2017

Expand abstractAutonomous robots can be difficult to design and understand. Designers have difficulty decoding the behaviour of their own robots simply by observing them. Naive users of robots similarly have difficulty deciphering robot behaviour simply through observation. In this paper we review relevant robot systems architecture, design, and transparency literature, and report on a programme of research to investigate practical approaches to improve robot transparency.We report on the investigation of real-time graphical and vocalised outputs as a means for both designers and end users to gain a better mental model of the internal state and decision making processes taking place within a robot. This approach, combined with a graphical approach to behaviour design, offers improved transparency for robot designers. We also report on studies of users’ understanding, where significant improvement has been achieved using both graphical and vocalisation transparency approaches.

University of Bath OPUS

ABOD3: A Graphical Visualization and Real-Time Debugging Tool for BOD Agents

and Andreas Theodorou
EUCog 2017 2016

Expand abstract

What Does the Robot Think? Transparency as a Fundamental Design Requirement for Intelligent Systems

Robert R. Wortham, Andreas Theodorou, and Joanna J. Bryson
In IJCAI-2016 Ethics for Artificial Intelligence Workshop 2016

Expand abstractAs robot reasoning becomes more complex, debugging becomes increasingly hard based solely on observable behaviour, even for robot designers and technical specialists. Similarly, nonspecialist users find it hard to create useful mental models of robot reasoning solely from observed behaviour. The EPSRC Principles of Robotics mandate that our artefacts should be transparent, but what does this mean in practice, and how does transparency affect both trust and utility? We investigate this relationship in the literature and find it to be complex, particularly in non industrial environments where transparency may have a wider range of effects on trust and utility depending on the application and purpose of the robot. We outline our programme of research to support our assertion that it is nevertheless possible to create transparent agents that are emotionally engaging despite having a transparent machine nature.

BibTeX, University of Bath OPUS

RobotTransparency,Trust and Utility

Robert R. Wortham, Andreas Theodorou, and Joanna J. Bryson
In AISB 16: EPSRC Principles of Robotics Workshop 2016

Expand abstractThe EPSRC’s Principles of Robotics dictates the implementation of transparency in robotic systems, however, research related to it is in its infancy. The current paper introduces the reader to the need of having transparent to inspection intelligent agents. We provide a robust definition of transparency, as a mechanism to expose the decision making of the robot, by considering and expanding upon other prominent definitions found in literature. The paper concludes by addressing potentials design decisions developers need to consider when designing transparent systems.

BibTeX, University of Bath OPUS

Why is my robot behaving like that? Designing transparency for real time inspection of autonomous robots

Andreas Theodorou, Robert R. Wortham, and Joanna J. Bryson
In AISB 16: EPSRC Principles of Robotics Workshop 2016

Expand abstractThe EPSRC’s Principles of Robotics dictates the implementation of transparency in robotic systems, however, research related to it is in its infancy. The current paper introduces the reader to the need of having transparent to inspection intelligent agents. We provide a robust definition of transparency, as a mechanism to expose the decision making of the robot, by considering and expanding upon other prominent definitions found in literature. The paper concludes by addressing potentials design decisions developers need to consider when designing transparent systems.

BibTeX, University of Bath OPUS

Projects

The Sustainability Game

Thumbnail of The Sustainability Game

The project aims at the development of a video game and then to explore if expertise in this game can help players to understand how to reach an optimal contribution to their group's public good or if it increases individually-oriented competition. We will test the hypothesis that playing games can help us find collaborative equilibria.

 

IEEE Global Initiative for Ethical Considerations in the Design of Autonomous Systems

Participating in the Initiative as a member of the P7001 standard, regarding transparency.

Homepage of the Initiative 

ABOD3

Thumbnail of ABOD3

ABOD3 is an integrated development environment for Behavior Oriented Design (BOD), providing new functionalities such as our novel approach to AI debugging. Please, consult GitHub or relevant publications for additional information.

GitHub Repository 

Teaching

2016/17

CM50230 and CM30229 Intelligent Control and Cognitive Systems
Lead Tutor

CM50258 Principles of Programming
Tutor / Marking Coordinator

CM10227 Programming 1A
Tutor / Marking Coordinator

CM10228 Programming 1B
Tutor / Marking Coordinator

2015/16

Programming 1A
Tutor

Programming 1B
Tutor / Marking Coordinator

2014/15

COMM051 Information Systems and Databases Development
Lab Demonstrator

COM2028 Artificial Intelligence
Lab Demonstrator

COM1029 Data Structures and Algorithms
Lab Demonstrator

Contact

E-Mail:

a.theodorou@bath.ac.uk

Other: andreas.theodorou@cs.gatech.edu

Drop Me a Line:

To a.theodorou@bath.ac.uk via Formspree.