Leveraging AI based Misbehaviour Detection for risk indicators in REWIRE project

by Suite5 Data Intelligence Solutions

In our rapidly evolving digital world, the REWIRE project stands out by developing a holistic framework for continuous security assessment and management of IoT devices throughout their entire lifecycle. Among the innovative technological artifacts offered by the REWIRE framework, stands the AI-based Misbehaviour Detection Engine, utilising advanced AI algorithms to identify and address misbehaviour. Hence, REWIRE is not only focusing on providing defence mechanisms applied to the lowest level of the Software/Hardware architecture stack (e.g. through the novel REWIRE attestation mechanisms), but focuses also to the application layer, applying innovative solutions to detect abnormal operational cases. REWIRE’s application of AI based misbehaviour detection is not just theoretical but grounded in two real-world use cases: Smart Satellites and Smart Cities. Each presenting unique challenges and opportunities for employing AI to enhance operational integrity and security measures; with the primary focus being on detecting potential misbehaviour and generating risks indicators.

Smart Satellites: Ensuring satellites’ operation

Considering that satellites are important for global communication and Earth observation, but also vulnerable to cyberattacks, or experience of implications (misbehaviours); REWIRE aims to make their operation more secure. Traditional security methods might not catch all the potential problems, so REWIRE will utilise its AI-based Misbehaviour Detection Engine to keep an eye on satellites’ operation, flag any abnormal behaviour and provide early warnings in the event of such abnormalities/misbehaviour.

The engine focuses on analysing variable satellite telemetry and employing advanced AI models, towards identifying unusual patterns in their operation that might suggest abnormal operation, addressing issues ranging from unexpected resets due to geomagnetic storms to anomalous battery charge levels. Among the various misbehaviour/plausibility checks that REWIRE is investigating, the ultimate focus is the design of an AI model that will be in position to generate early warnings by analysing the behavioural patterns of satellites and the geomagnetic activity in order to support the decision making of satellite operators. This approach not only aims to improve the safety of satellites but also provide additional means for making space critical infrastructure more secure also enhancing their reliability and ensuring missions’ continuity.

Smart Cities: Enhancing Risk detection

In the context of Smart Cities, REWIRE extends its expertise in AI-based misbehaviour detection to proactively identify risks in urban environments with special attention to safety critical applications. As cities become more digitised, ensuring the security of their complex systems becomes paramount; here REWIRE aims to extend its AI-based misbehaviour detection competency to safe critical tasks such as pedestrian traffic control management, fire suppression systems control, etc.  As such, REWIRE will utilise AI to detect misbehaviour or abnormalities in safe critical operations undertaken in the context of smart cities, serving as an effective risk mitigation tool. The early detection of anomalies ensures that potential risks are flagged, allowing for timely intervention, and preventing adverse outcomes.

REWIRE’s implementation of AI-based misbehaviour detection goes beyond ensuring ethical usage; it actively provides risk indicators when abnormalities are detected. In satellite technology and Smart Cities, this approach offers a proactive attitude in managing potential risks.  In conclusion, the REWIRE project exemplifies the crucial role of AI in modern security effort; contributing to the creation of robust systems that can promptly address and mitigate risks, ensuring the reliability and security of space and urban infrastructures technology, paving the way for a more secure future in our interconnected world.

Leave a Reply