The Trusted Autonomous Systems Hub pulls researchers from computer science and engineering together to develop the trustable autonomous systems of tomorrow. Our vision for such autonomous systems is that they are capable of reasoning and planning, they are safe and secure, they efficiently integrate in human-autonomous systems teams, they rely on wireless communications and they might have physical embodiment as robots or intelligent sensors, they interact with humans, they are accountable for their behaviour, hereby allowing users to place their trust in them.
Research Questions
The objective of the Trusted Autonomous Systems hub is to advance the state-of-the-art in trustworthy autonomous systems and accountable human-autonomy teaming. The multi-disciplinary team of computer scientists and engineers investigates key research questions, including:
- What are the technical challenges involved in an effective and reliable autonomy?
- What are the technical challenges involved in creating trustable systems for decision making in human-robotic teaming?
- What are the technical challenges in creating multi-robot teams and designing human-robot interaction?
- What are the key factors, opportunities and risks for robotics and autonomous systems operating in challenging environments?
People
Senior Lecturer in AI Education
Lecturer in Robotics and Autonomous Systems
Lecturer in Autonomous Systems
Senior Lecturer in Computer Science Education
Senior Lecturer
Professor of Computer Science
Projects
Persistent Autonomy in Underwater Missions.
The KCL Pandora team consists of Professors Maria Fox and Derek Long, working with Research Fellow Dr Daniele Magazzeni. The three researchers have worked together for some time, focussing on continuous planning and hybrid control systems. They devised a plan-based policy learning approach described in their ICAPS paper in 2011, awarded best research paper prize. Subsequently, the work has been applied to the problem of bloom following, described in another ICAPS paper, in 2012. Professor Fox is leading WP2, exploiting planning for the intelligent control of AUVs in persistent autonomous missions. The team is also contributing to various aspects of the project, integrating planning into the management of learned behaviours and the links from sensor data to symbolic representations of the world state.
ROSPlan: Artificial Intelligence Planning for Robots
The ROSPlan framework provides a collection of tools for AI Planning in a ROS system. ROSPlan has a variety of nodes which encapsulate planning, problem generation, and plan execution. It possesses a simple interface, and links to common ROS libraries.
H2020 ERGO: Robotic Goal-Oriented Autonomous Controller
An H2020 project funded by the European Commission within Strategic Research Cluster on Space Robotics Technologies
H2020 SQUIRREL: Autonomous Robots deployed in unstructured domestic settings
Clutter in an open world is a challenge for many aspects of robotic systems, especially for autonomous robots deployed in unstructured domestic settings, affecting navigation, manipulation, vision, human robot interaction and planning. SQUIRREL addresses these issues by actively controlling clutter and incrementally learning to extend the robot’s capabilities while doing so. We term this the B3 (bit by bit) approach, as the robot tackles clutter one bit at a time and also extends its knowledge continuously as new bits of information become available. SQUIRREL is inspired by a user driven scenario, that exhibits all the rich complexity required to convincingly drive research, but allows tractable solutions with high potential for exploitation. We propose a toy cleaning scenario, where a robot learns to collect toys scattered in loose clumps or tangled heaps on the floor in a child’s room, and to stow them in designated target locations.
People
Senior Lecturer in AI Education
Lecturer in Robotics and Autonomous Systems
Lecturer in Autonomous Systems
Senior Lecturer in Computer Science Education
Senior Lecturer
Professor of Computer Science
Projects
Persistent Autonomy in Underwater Missions.
The KCL Pandora team consists of Professors Maria Fox and Derek Long, working with Research Fellow Dr Daniele Magazzeni. The three researchers have worked together for some time, focussing on continuous planning and hybrid control systems. They devised a plan-based policy learning approach described in their ICAPS paper in 2011, awarded best research paper prize. Subsequently, the work has been applied to the problem of bloom following, described in another ICAPS paper, in 2012. Professor Fox is leading WP2, exploiting planning for the intelligent control of AUVs in persistent autonomous missions. The team is also contributing to various aspects of the project, integrating planning into the management of learned behaviours and the links from sensor data to symbolic representations of the world state.
ROSPlan: Artificial Intelligence Planning for Robots
The ROSPlan framework provides a collection of tools for AI Planning in a ROS system. ROSPlan has a variety of nodes which encapsulate planning, problem generation, and plan execution. It possesses a simple interface, and links to common ROS libraries.
H2020 ERGO: Robotic Goal-Oriented Autonomous Controller
An H2020 project funded by the European Commission within Strategic Research Cluster on Space Robotics Technologies
H2020 SQUIRREL: Autonomous Robots deployed in unstructured domestic settings
Clutter in an open world is a challenge for many aspects of robotic systems, especially for autonomous robots deployed in unstructured domestic settings, affecting navigation, manipulation, vision, human robot interaction and planning. SQUIRREL addresses these issues by actively controlling clutter and incrementally learning to extend the robot’s capabilities while doing so. We term this the B3 (bit by bit) approach, as the robot tackles clutter one bit at a time and also extends its knowledge continuously as new bits of information become available. SQUIRREL is inspired by a user driven scenario, that exhibits all the rich complexity required to convincingly drive research, but allows tractable solutions with high potential for exploitation. We propose a toy cleaning scenario, where a robot learns to collect toys scattered in loose clumps or tangled heaps on the floor in a child’s room, and to stow them in designated target locations.
Group leads
Professor of Computer Science
Senior Lecturer in Artificial Intelligence