Recipient organisationCoventry UniversitySource-published name: Coventry University
Funding£29K
PeriodMar 2025 — Jul 2025
In plain English
AI plain-English summary
A single modern vehicle can take over 300 hours of manual cybersecurity analysis per electronic control unit, and there are not enough experts to keep up. AutoTARA replaces much of that manual work with an AI-driven platform that automates threat identification, risk assessment, and compliance checks for automotive cybersecurity. The problem is acute: modern cars run on software, receive over-the-air updates, and connect to networks, making them vulnerable to cyberattacks. The industry is required to follow ISO/SAE 21434, a standard that mandates Threat Analysis and Risk Assessment (TARA), but doing this manually is slow, inconsistent, and expensive. AutoTARA uses large language models, attack graphs, and probabilistic risk engines to automate reasoning about vulnerabilities, cutting execution costs by 60% and accelerating certification timelines. If successful, the platform—commercialised as a cloud-based or on-premises SaaS tool—would let manufacturers, suppliers, and consultancies integrate faster, more consistent cybersecurity checks into vehicle development. It also aligns with emerging regulations such as UNECE R155 and the Cyber Resilience Act, helping the industry meet compliance requirements without relying on a scarce workforce of cybersecurity specialists.
View original technical description
The AutoTARA project presents a transformative solution to one of the automotive industry's most pressing challenges: ensuring robust, scalable, and cost-effective cybersecurity in an era of increasing vehicle connectivity and digitalization. As modern vehicles incorporate software-defined architectures and over-the-air updates along with high connectivity and complexity, they become more vulnerable to cyber threats. The industry faces a severe shortage of cybersecurity experts, leading to inconsistencies and delays in Threat Analysis and Risk Assessment (TARA), a mandatory process under ISO/SAE 21434 for regulatory compliance. Traditional manual TARA methods are time-consuming, costly, and error-prone, taking over 300 hours per ECU, and limiting scalability as automotive systems grow in complexity. AutoTARA will revolutionise this process by integrating advanced AI models, machine learning pipelines, and cybersecurity-specific engines into an intelligent, semi-automated assessment framework. Unlike existing tools that rely on manual threat modelling, AutoTARA employs natural language processing (NLP), attack graph-based threat modelling, and probabilistic risk assessment to analyze cybersecurity assets, system architectures, and historical vulnerability databases. The platform automates threat identification, risk quantification, and compliance validation, ensuring faster, more consistent, and scalable cybersecurity risk assessments. What differentiates AutoTARA is its multi-layered AI architecture, which combines large language models (LLMs), in-house AI threat intelligence engines, and rule-based cybersecurity frameworks. This hybrid approach enables automated reasoning, reducing reliance on scarce cybersecurity professionals while enhancing accuracy and adaptability. Furthermore, AutoTARA aligns with emerging cybersecurity regulations, including UNECE R155 and the Cyber Resilience Act (CRA), ensuring compliance and regulatory readiness for vehicle manufacturers. It aimed to be commercialised as a SaaS platform, targeting OEMs, Tier-1 suppliers, and cybersecurity consultancies. It offers cloud-based and on-premises deployment, allowing organizations to seamlessly integrate AI-driven TARA into their cybersecurity workflows. By reducing execution costs by 60% and accelerating certification timelines, AutoTARA establishes a new industry benchmark for automotive cybersecurity risk assessment. Led by Coventry University's Centre for Future Transport and Cities (CFTC) , the project benefits from deep expertise in automotive cybersecurity, validation, and verification. The team, accredited in ISO/SAE 21434 compliance and with extensive research experience in automotive security assessment, brings proven methodologies for threat analysis and formal verification of cybersecurity measures. Their contributions ensure rigorous testing, regulatory alignment, and scientific validation, making AutoTARA a trusted, high-impact solution for the evolving cybersecurity landscape of connected and modern vehicles.[][0] [0]: https://www.coventry.ac.uk/research/areas-of-research/centre-for-future-transport-and-cities/
Plain English summaries and category classifications on this site are generated by AI and may not perfectly reflect the original research.
Is something wrong? Let us know