Towards Adaptively Morphing Embedded Systems

Project Information

ADMORPH

Grant agreement ID: 871259

DOI

10.3030/871259

Project closed

EC signature date 5 November 2019

Start date 1 January 2020

End date 30 June 2023

Funded under

• INDUSTRIAL LEADERSHIP - Leadership in enabling and industrial technologies - Information and Communication Technologies (ICT)

Total cost

€ 4 499 468,75

EU contribution

€ 4 499 468,00

Coordinated by

UNIVERSITEIT VAN AMSTERDAM
Netherlands

Periodic Reporting for period 2 - ADMORPH (Towards Adaptively Morphing Embedded Systems)

Reporting period: 2021-11-01 to 2023-06-30

Summary of the context and overall objectives of the project

The aim of the ADMORPH project is to make various types of complex systems that are controlled by computers more resistant to defects and more secure. The project has received a grant of 4.5 million euros from the EU Horizon2020 programme. ADMORPH is an international European project led by the University of Amsterdam.

The world around us is increasingly monitored and controlled by computers. Smart algorithms control all sorts of factory processes and entire distribution chains, vehicles like cars and airplanes can now drive or fly without a human driver, and with the most modern surveillance systems it is no longer necessary that a person in a box keeps an eye on the monitors all night.

Safety
In technical terms, this phenomenon in which physical systems in daily live are controlled via computer technologies are called Cyber Physical Systems (CPS). For CPS, to function properly, it is important that they are resistant to hardware defects and are safe; and the more complex the system, the greater the challenge. Many CPS have applications where reliability is very important. Think of the autonomous cars and airplanes. The systems also sometimes operate in harsh outdoor conditions, so they must therefore be able to deal with system components that become abruptly defective. And of course all systems must be well protected against hackers. Otherwise you run the risk that the factory processes and distribution chains will stop, vehicles will get into accidents and surveillance systems will no longer pick up suspicious situations.

European collaboration
Making complex CPS more robust and secure is the goal of the ADMORPH consortium. ADMORPH is an international European project led by the University of Amsterdam (UvA). The other partners are Thales Nederland, SYSGO S.A.S. from France, the University of Luxembourg, Lund University from Sweden, the United Technologies Research Center Ireland, the Czech Q-media, the Faculty of Sciences of the University of Lisbon from Portugal and the University of Augsburg from Germany. The acronym ADMORPH stands for Adaptive, dynamically morphing, mission and safety-critical CPS. As the name suggests, the parties involved have devised a strategy to make the systems more robust and secure by making them more adaptive and flexible.

Conclusion
After the project duration of three years, the team has not only worked out a set of practical solutions for specifying, designing, analyzing and implementing adaptive CPS, but has also applied some of these solutions to a number of existing systems. These real-world applications the project focuses on are: a complex system that controls autonomous aircrafts, a system that provides real-time radar surveillance, and a system that monitors and manages a metro network.

Work performed from the beginning of the project to the end of the period covered by the report and main results achieved so far

During this 3-year project, the ADMORPH project team has studied and developed a range of different solutions for specifying, designing, analyzing and implementing adaptive CPS. This work has resulted in a ‘tool-box’ of methods and techniques as well as accompanying software-tools for various lifecycle-stages (specification, analysis, design, operation, maintenance, etc.) of adaptive CPS. More specifically, this tool-box consists of
• a coordination language, compiler and middleware for (specifying) adaptive CPS;
• a multi-model model-of-computation to formally model adaptive real-time embedded systems;
• a system simulator and design-space exploration tools for analyzing and optimizing the lifetime of adaptive computer systems;
• adaptive Byzantine fault-tolerance and rejuvenation analysis tools for adaptive
systems, including degraded modes;
• analysis tools for fault-resilient control systems;
• run-time environments for fault-detection, resilient control and task re-execution;
• software update framework without loss of service.

The application of these methods, techniques and tools has been demonstrated in the scope of seven demonstrators produced by the ADMORPH consortium. Three of these demonstrators concern of existing, real-world systems, namely a complex system that controls autonomous aircrafts, a system that provides real-time radar surveillance, and a system that monitors and manages a metro network.

The ADMORPH results have been disseminated using a variety of different channels. For example, there is an active LinkedIn community around ADMORPH (the project has 223 followers on LinkedIn) and the ADMORPH website has been visited frequently. Moreover, the ADMORPH consortium members organized a number of events (like special sessions and workshops) at, for example, the HiPEAC, DATE and RTSS conferences. Last but not least, the project produced a substantial number of scientific research papers. Regarding the exploitation of research results, various activities have been initiated. For instance, a so-called TechTracker (overview of open-source software tools developed in ADMORPH) and DemoTracker (overview of developed demo’s) were produced. Furthermore, as part of NextGenCPSoS, which is a cluster of seven EU funded projects (CPSoSaware, SMART4ALL, ADEPTNESS, ADMORPH, HiPEAC, DIH4CPS and TEACHING), aiming at developing new technologies and approaches for cyber-physical systems of systems, our project participated in creating an informative factsheet and video.

Progress beyond the state of the art and expected potential impact (including the socio-economic impact and the wider societal implications of the project so far)

Deploying system adaptivity techniques for making computer systems more robust against components failures or cyber-attacks is a topic that already received the necessary research attention. However, there exists no systematic design methodology or comprehensive design framework for adaptive CPS that aims at achieving verifiable system robustness against both hardware component failures and cyber-attacks while respecting time and energy constraints and keeping defined quality-of-service levels up as long as possible. ADMORPH aims at providing a breakthrough towards that end by developing a novel, holistic approach to the specification, design, analysis and runtime deployment of mission- and safety-critical adaptive embedded computer systems in CPS. By ensuring direct industrial relevance, this could, eventually, lead to the first industry-grade design framework for adaptive, robust (mission- and safety-critical) computer systems.

ADMORPH project logo