Intelligent Transportation Systems (ITS) are becoming dependent on connected vehicles, roadside infrastructure, IoT devices, communication networks, and intelligent computing platforms, which introduces major reliability and cybersecurity challenges. The present paper presents a co-design framework for predictive maintenance and cyber security based on artificial intelligence (AI) for improving the cyber-physical resilience in ITS. The framework will combine equipment health, maintainability data, operational sensor data, and network-security data to analyze both the physical degradation and cyber threats. A hybrid model of XGBoost-BiLSTM-Attention is designed, in which XGBoost picks informative features, BiLSTM models the temporal dependency and attention focuses on important observations. The framework enables the detection of the health of the equipment, the detection of anomalies, failure prediction, remaining useful life estimation, intrusion detection and identification of cyber-attacks. There is a joint classification mechanism which separates normal operation, physical faults, cyber-attacks and simultaneous fault–attack conditions. The coordinated risk assessment is enabled by Unified risk assessment. The proposed approach helps to mitigate false alarms, enhance fault–attack discrimination, enable proactive maintenance and boost the reliability, security, and resiliency of next generation ITS infrastructures.