Skip to main content

Work Package 2 – Architectures

Designing flexible and transparent software architectures to enable autonomous robots to be deployed reliably and verifiably in demanding and long-lasting scenarios.

A software architecture is a collection of software elements, together with the relationships between them. In robotic systems these elements are often associated with modular components that, together, describe and implement the overall behaviour of the system. WP2 will explore modular, transparent, resilient, and verifiable architectures applicable across sectors, targeting both individual and interacting systems, and providing both strong behavioural guarantees and guidelines for their use.

ARCHITECTURAL TRANSPARENCY

Not only are architectures important in highlighting structural and design decisions, but the way we build these architectures has a fundamental impact on the effectiveness of our systems. Building them in the right way can promote transparency (hence: explainability; verification; assurance), flexibility (to cope with unforeseen issues), and resilience (reliability and recovery). Building them carelessly can make all these aspects difficult (and sometimes impossible). We aim to provide guidance on, and patterns for, best practice.

SYSTEM AWARE ARCHITECTURES

A key aspect of developing robots for demanding and long-lasting environments will be their resilience, especially their ability to cope with unexpected situations.  It is here that autonomous decisions will need to be made. This will require software reconfigurability and an awareness, in the robot itself, about what architectural changes are possible and how those changes will affect the capabilities of the robot. Ensuring effective and appropriate architectural mechanisms to achieve this “self-awareness” is an important target.

MECHANISMS FOR ROBOT COLLECTIVES

Architectural aspects are important, not only for individual robots, but for both multi-robot and human-robotic interacting systems. In collaboration with WP3 we aim to explore these aspects, providing reliable and transparent approaches to architectures for both swarms and human-robot teams.

Updates

Research Spotlight – Dr Raynaldio Limarga

Ray’s work will involve examining and developing the architectures needed for resilient and trustworthy autonomous systems while also ensuring that the core autonomous decision-making is both identifiable and verifiable. This involves empowering the systems to navigate and respond effectively to unforeseen or unpredictable scenarios.  The team will build resilience into every design layer while maintaining transparency and accountability, ensuring every decision is traceable and justifiable. Only by this can we establish strong trust in the overall decision-making processes. Trust in autonomous systems is paramount to ensure engineers, users, and regulators feel confident in deploying robotics in complex and dynamic environments.

Research Spotlight – Dr Sen Zheng

Sen’s research will focus on the analysis of various types of robot collectives, including human-robot teams, robot teams, and robot swarms. Unlike single robots, collectives offer greater fault tolerance, enabling them to operate efficiently and withstand faults even during performance degradation, making them highly promising for industrial applications, such as environmental protection, disaster response, nuclear, and space missions. This work, between WP2 and WP3 aims to develop mathematically rigorous methodologies and practical tools to ensure the reliability and transparency of robot collectives, instilling greater confidence in stakeholders throughout the development and deployment of these systems.

    Explore Work Packages

    WP 1 Components
    WP2 Architectures
    WP3 Interactions
    WP4 Assurance
    WP5 Demonstrators