Tele-operated urban delivery vehicle which enables unmanned last mile delivery operation for safe, cost effective and all-day operation
Teleoperated last-mile delivery vehicles provide a flexible and efficient solution for transporting goods in both urban and rural environments. They represent an important transitional step from driver-assisted vehicles to fully autonomous vehicles.
By combining remote control with AI-assisted perception and decision support, these systems address challenges related to safety, functionality, and cost. However, their reliance on connectivity, real-time data processing, and AI-driven features introduces significant cybersecurity risks.
This use case focuses on securing AI-enabled connectivity, driver assistance, and teleoperation systems in teleoperated delivery vehicles.
Security challenges
As a high-risk, connected AI system operating in public environments, this technology faces multiple security challenges:
- Protection of AI models against manipulation or adversarial attacks
- Securing 4G/5G and wireless communication channels
- Ensuring integrity of over-the-air (OTA) updates
- Preventing unauthorized access to teleoperation controls
- Maintaining system availability and resilience in real time
A successful attack could compromise vehicle control, disrupt operations, or endanger public safety.
SHASAI’s Contribution
SHASAI strengthens this teleoperated vehicle system by:
- Applying secure-by-design principles across AI, software, and hardware components
- Assessing vulnerabilities in AI-driven perception and control modules
- Securing communication and OTA update mechanisms
- Enabling automated testing against AI-specific and infrastructure-level threats
- Supporting continuous monitoring and anomaly detection during operation
Through this validation scenario, SHASAI demonstrates how secure AI methodologies can enhance resilience, safety, and regulatory compliance in connected and semi-autonomous mobility systems.