ISO / SAE Standard 21434 sets strict engineering requirements for the cybersecurity risk management throughout the entire lifecycle of connected vehicles, covering stages from initial concept and software/hardware design to production, post-production operations, and decommissioning. Tensor Obtaining certification with zero non-compliances proves that its cybersecurity management system CSMS is in line with the cybersecurity management requirements stipulated by United Nations Regulation No. 155 UN R155. UN R155 This standard applies to vehicle cybersecurity type certifications in over 60 countries and regions, including the European Union, the United Kingdom, Japan, and South Korea.

Tensor Founder and CEO Dr. Jay Xiao means:

In the Physical AI era, network security is not just about protecting software; it's also about safeguarding individual autonomy. Traditional automobile manufacturers produce connected cars that continuously transmit data back to corporate clouds, whereas the design goal of Tensor Robocar is to be solely responsible to its owners. Completing the ISO / SAE 21434 audits with no findings indicates that our security design process meets the standards expected by global regulatory authorities, and it also demonstrates that decentralized, in-vehicle AI supercomputing can integrate network security from the very beginning.

Zero Trust Security for Edge Devices Physical AI

Tensor indicates that, unlike the B2B commercial robot taxi fleet or traditional electric vehicle architectures that rely on the cloud, Tensor Robocar runs complex environmental perception, trajectory planning, and biometric authentication locally on the vehicle, utilizing its high-performance in-vehicle supercomputing stack. The review process involving Applus + IDIADA covers the network security procedures that support the Tensor edge-first architecture, which is designed to operate without the need for a continuous cloud connection.

The core engineering processes that are subject to review include:

• Full Lifecycle Security Engineering: By requiring that each piece of network security work have a unique identifier and version control, and by tracing from the top-level objectives of the vehicle back to component requirements and verification tests, traceability is established throughout all stages of vehicle development.
• Strict Threat Analysis and Risk Assessment ( TARA ): By employing the ISO / SAE 21434 method, TARA is carried out at both the vehicle and ECU levels. A traceable relationship is established between assets, threat scenarios, attack paths, risks, and control measures to demonstrate systematic threat coverage.
• Vehicle Safety Operation Center ( VSOC ): An integrated vehicle safety operation center has been deployed for the vehicles in use, featuring intrusion detection, hierarchical event escalation, and clear response time limits.
• Secure Over-the-Air Download ( OTA ) Update: Utilize Tensor's proprietary OTA platform for managing, deploying, and monitoring firmware packages, and verify the end-to-end integrity of software delivery through comprehensive encryption controls.
• Multi-layer defense verification: By conducting comprehensive penetration testing, fuzz testing, communication security testing, and static/dynamic code analysis, the actual anti-attack capabilities are confirmed, and these tests are directly linked to the requirements generated by TARA.
• Supply Chain Cybersecurity Governance: A structured assessment framework is adopted for suppliers related to cybersecurity, covering governance, secure development practices, vulnerability management, incident support, evidence delivery, and change control.
• Production and final inspection verification: Directly extend network security requirements to vehicle flashing, configuration, functional verification, and final inspection testing to prevent vulnerabilities from being introduced during the manufacturing process, and complete the checks before commercial release.

Applus + IDIADA 的 Alejandro Manilla represents:

"Our assessment conducted an independent review of Tensor's network security processes and provided an evaluation that supports structured automotive network security methods in accordance with the requirements of ISO / SAE 21434. Tensor has proven to meet the applicable requirements of these standards, and we look forward to continuing our cooperation in future initiatives."

Promote global expansion and technology licensing

Tensor indicates that the ISO / SAE 21434 certification is an important milestone on its path to the global market, laying the foundation for vehicle certification and UN R155 type approval in Europe, the UK, South Korea, Japan, and other regions. Tensor continues to focus on providing fully autonomous capabilities, while also ensuring strict digital and physical security.

In addition to Tensor Robocar, Tensor has also developed an authorized full-stack AI driving technology platform for vehicle manufacturers and first-tier suppliers. The company states that ISO / SAE 21434 certification has strengthened the network security foundation of this platform, proving that the processes supporting its vehicle architecture, software, in-vehicle computing, security OTA updates, and vehicle safety operations meet the highest standards.

About Tensor

Tensor is an American Physical AI technology company that develops AI with agent capabilities and full-stack autonomous driving technology. Its flagship product, Tensor Robocar, is the world's first automotive-grade SAE L4 driverless vehicle, which can be owned privately or used for robot taxi deployment. In addition to Robocar, Tensor also licenses its full-stack Physical AI platform to automakers and tier-one suppliers through Autonomous Driving and as in a Service ( ADaaS ) model.

Driven by the intelligent agent AI, which possesses reasoning capabilities, Robocar demonstrates how Physical AI can change the way people interact with machines. Robocar is designed to give users the freedom to choose between driving and being driven, featuring a foldable steering wheel, retractable pedals, and a sliding display screen. Tensor was founded in Silicon Valley in 2016 and currently has offices in San Jose, Barcelona, Singapore, and Dubai.

Regarding Applus + IDIADA

Applus + IDIADA is a global partner in the automotive industry, providing design, engineering, testing, and certification services. IDIADA has over 30 years of experience and its business covers 24 countries, offering comprehensive technical support for international vehicle certification and regulatory compliance.