Abstract: Systems and methods for detecting and remediating roadway hazards are disclosed. A machine learning model is trained on a dataset related to roadway items. Input data is collected by a data collection engine and provided to a pattern recognizer. The pattern recognizer extracts roadway features and recognized patterns from the input data and provide the extracted features to a trained machine learning model. The trained model compares the extracted features to the model, and a risk value is generated. The risk value is compared to a risk value threshold. If the risk value is equal to or exceeds the risk threshold, then the input data may be classified as a roadway hazard. Remedial action is subsequently be triggered.

Abstract: Examples of the present disclosure describe systems and methods for detecting and remediating roadway hazards. In example aspects, a machine learning model is trained on a dataset related to roadway items. Input data may then be collected by a data collection engine and provided to a pattern recognizer. The pattern recognizer may extract roadway features (physical and non-physical) and recognized patterns from the input data and provide the extracted features to a trained machine learning model. The trained machine learning model may compare the extracted features to the model, and a risk value may be generated. The risk value may be compared to a risk value threshold. If the risk value is equal to or exceeds the risk threshold, then the input data may be classified as a roadway hazard. Remedial action may subsequently be triggered, e.g., notifying other vehicles on the roadway, notifying other devices of drivers on the roadway, and/or notifying interested third parties (e.g.

Abstract: Examples of the present disclosure describe systems and methods for detecting and remediating roadway hazards. In example aspects, a machine learning model is trained on a dataset related to roadway items. Input data may then be collected by a data collection engine and provided to a pattern recognizer. The pattern recognizer may extract roadway features (physical and non-physical) and recognized patterns from the input data and provide the extracted features to a trained machine learning model. The trained machine learning model may compare the extracted features to the model, and a risk value may be generated. The risk value may be compared to a risk value threshold. If the risk value is equal to or exceeds the risk threshold, then the input data may be classified as a roadway hazard. Remedial action may subsequently be triggered, e.g., notifying other vehicles on the roadway, notifying other devices of drivers on the roadway, and/or notifying interested third parties (e.g.