Connected Autonomous Vehicle (CAV) platoons have been extensively studied to protect against cyber and physical vulnerabilities. Faults can occur in all layers of the platoon system or could be introduced by impaired human drivers. Since different types of faults may require different fault resolution methods, identifying the fault class facilitates the selection of the best mitigation strategy. This paper introduces a Multi-Head Attention Machine Learning (MHA-ML) approach to classify a set of five different faults and abnormalities in mixed autonomous and human-driven vehicle platoons. Autonomous vehicles can face actuator faults, False Data Injection (FDI) attacks, and Denial-of-Service (DoS) attacks, while abnormalities such as drunk or distracted human drivers could occur. MHA-ML is developed to identify faulty vehicle behavior over long sequences of sensor measurements. MHA-ML is trained on a mixed platoon simulation model and then tested on mobile laboratory robots. The experiment classifies the five fault categories with 90% accuracy and outperforms a baseline recurrent neural network approach.

This study proposes an algorithm for fault detection and mitigation of mixed autonomous and human-driven vehicle platoons based on transmissibility identification. This work is motivated by the fact that on-road human-drivers’ behaviour is unknown and difficult to be predicted. Transmissibility operators are mathematical operators that relate one subset of outputs to another in the same system. The transmissibility superiority is represented in the in-dependency on the system excitation signals. We reformulate the system dynamics to render the system inputs, external disturbances, as well as the human-drivers’ behaviour along with any other nonlinearities as independent excitation signals on the system. Therefore, the transmissibility operators become independent of the human-drivers’ behaviour and robust against external disturbances. Transmissibilities are then applied to detect and localize physical and cyber faults within the platoon. Then these faults are mitigated using a transmissibility-based sliding mode controller. The controller stability and the string stability are investigated while the controller is active and the faults are mitigated. We validate the proposed algorithm on a model of the platoon obtained using the bond graph approach. Moreover, we apply the proposed algorithm experimentally to a platoon consisting of three robots (i.e. two autonomous robots and a human-driven robot), that is connected using wireless communications.Note to Practitioners—The existence of connected autonomous vehicles depends greatly on the smooth transition between the current on-road human-driven vehicles to autonomous vehicles. The typical methods of securing dynamic systems depend on estimating the system behaviour and responses. Increasing the number of autonomous vehicles on roads necessitates the typical securing techniques to estimate the human-drivers’ behaviour. Thus, securing the connected autonomous vehicles during this transition is challenging since the on-road human-driver behavior is unknown and difficult to be estimated. Moreover, connected autonomous vehicles should adapt to their environment while maintaining their role within the autonomous platoon. This adaptation includes adapting to the unknown human-driver behaviour. This inspired the authors to develop the proposed transmissibility-based fault mitigation. The proposed technique is shown to be able to handle unknown human-driver behaviors, different driving conditions such as road irregularities and different weather conditions, and different physical and cyber faults (i.e., in the vehicles or in the communication links). The platoon stability is then investigated while the faults are mitigated, and shown to guarantee the platoon stability.

Fault mitigation in Connected Autonomous Vehicle (CAV) platoons is faster and more reliable if the fault structure is known. In this paper we propose using transmissibility operators, which are relationships that relate a set of velocities with another in the platoon, to classify the faults. Transmissibility operators were shown to be exceptional in signals estimation; however, its also shown to be noncausal and thus can only be used offline. To this end, we propose using Data Aggregation (DAgger), which is an extension in imitation learning to transfer the classification experience from transmissibility operators to a novice machine learning agent to be used online. A heterogeneous CAV platoon was modeled with three different faults separately. These faults are actuator disturbances, false data injection attacks, and communication time delay. The proposed algorithm is then tested on the platoon model and then applied to an experimental setup that consists of three autonomous robots. The overall classification accuracy achieved was 95.8% for the experiment.

Different health-monitoring techniques were considered in the literature to enhance the safety and stability of Connected Autonomous Vehicle (CAV) platoons. The health-monitoring processes include fault detection, localization, and mitigation. It is evident that mitigating these faults is faster and more reliable if the fault structure is known. To this end, we consider classifying the fault class using supervised machine learning. We first model a heterogeneous CAV platoon with three different common faults separately. These faults are bounded actuator disturbances (namely, engine bearing knock), False Data Injection (FDI) attack, and communication time delay. We consider two supervised machine learning classifiers, the first classifier determines whether the fault is bounded disturbances or communication delay, and the second classifier determines whether the disturbances are in the physical or cyber layer. We have compared four machine learning techniques for each classifier, Support Vector Machine (SVM), Naive Bayes (NB), Quadratic Discriminant (QD), and K-Nearest Neighbors (KNN). The classifiers are trained firstly on the simulation model, then are tested on a different set of observations and tested experimentally on a platoon of three autonomous robots. The highest accuracy was achieved by considering SVM for the first classifier and QD for the second classifier. The overall classification accuracy achieved is 96.8% for the simulation test and 92.1% for the experiment.

An autonomous vehicle platoon is a network of autonomous vehicles that communicate together to move in a desired way. One of the greatest threats to the operation of an autonomous vehicle platoon is the failure of either a physical component of a vehicle or a communication link between two vehicles. This failure affects the safety and stability of the autonomous vehicle platoon. Transmissibility-based health monitoring uses available sensor measurements for fault detection under unknown excitation and unknown dynamics of the network. After a fault is detected, a sliding mode controller is used to mitigate the fault. Different fault scenarios are considered including vehicle internal disturbances, cyber attacks, and communication delays. We apply the proposed approach to a bond graph model of the platoon and an experimental setup consisting of three autonomous robots.