Landslide characteristics and its impact on tourism for two roadside towns along the Kathmandu Kyirong Highway

Frequent landslide events affect the Kathmandu Kyirong Highway(KKH),one of the most strategic Sino-Nepal highways,with multiple social effects.Amongst them,the impacts on local tourism,although being substantial,have not been studied so far.The aim of this research is to analyze the characteristics of such landslides and their influence on road damages and/or blockages as well as on local tourism industry.We analyzed the co-seismic landslides triggered by the Gorkha Earthquake,2015(7.8 Mw),the post-seismic landslides that occurred during the monsoons following the earthquake,as well as landslides which occurred or reactivated in 2018,with relation to the damage that they caused to the highway.High resolution satellite images from 2015 to 2018,and field data were used for the analysis.The Langtang avalanche that locates off the highway was also mapped due to its high impacts on tourism.Between 2015 and 2018,the number of road damaging landslides in the Betrawati-Rasuwagadhi section of KKH(where Dhunche and Syafrubesi towns are located)was 101 in the main track(MT)and 103 in the new track(NT),with respective average density of 1.46/km and 3.63/km.The dominant observed landslide types were debris slides and rock falls.Landslides were mostly concentrated in the locations with the following characteristics:1)having higher elevated area,2)being located with the‘main central thrust’and other lineaments’belts,3)belonging to the Proterozoic lesser Himalayan rocks,4)having a slope gradient of 25°-45°,5)having northern,western and southern slope aspect,6)being subjected to average annual rainfall of higher than 1,000 mm,and 7)having less than 4 km distance from the past earthquake epicenters.The results further indicated that 7 rain-induced and 4 co-and post-seismic landslides have great impact on tourist flows.An impact analysis was also assessed through a door to door questionnaire survey with local hotel operators from Dhunche and Syafrubesi towns(n=29+31).The results reveal that out of six rigorously affected sectors by landslides leading to road blockage,tourism business is the most impacted livelihood sector in these towns.The reduction of visitors in different hotels ranged from 50%-100%in Dhunche and 70%-100%in Syafrubesi for the first year aftermath of the tremor.This is higher than the respective 5%-50%tourist reduction due to raininduced landslides.Using as a reference the base year 2014,the income loss of hotels in both towns was found to be 50%-100%in 2015,20%-100%in 2016,5%-75%in 2017,and similar to 35%in 2018.These results provide insights on the synergic effect of contributing factors for cut slope as well as down slope instability along mountainous motorways and their impact on income sources for local communities.

Road Damage Detection and Classification Using Mask R-CNN with DenseNet Backbone

Automatic road damage detection using image processing is an important aspect of road maintenance.It is also a challenging problem due to the inhomogeneity of road damage and complicated background in the road images.In recent years,deep convolutional neural network based methods have been used to address the challenges of road damage detection and classification.In this paper,we propose a new approach to address those challenges.This approach uses densely connected convolution networks as the backbone of the Mask R-CNN to effectively extract image feature,a feature pyramid network for combining multiple scales features,a region proposal network to generate the road damage region,and a fully convolutional neural network to classify the road damage region and refine the region bounding box.This method can not only detect and classify the road damage,but also create a mask of the road damage.Experimental results show that the proposed approach can achieve better results compared with other existing methods.

A Deep Learning-Based Approach for Road Surface Damage Detection

Timely detection and elimination of damage in areas with excessive vehicle loading can reduce the risk of road accidents.Currently,various methods of photo and video surveillance are used to monitor the condition of the road surface.The manual approach to evaluation and analysis of the received data can take a protracted period of time.Thus,it is necessary to improve the procedures for inspection and assessment of the condition of control objects with the help of computer vision and deep learning techniques.In this paper,we propose a model based on Mask Region-based Convolutional Neural Network(Mask R-CNN)architecture for identifying defects of the road surface in the real-time mode.It shows the process of collecting and the features of the training samples and the deep neural network(DNN)training process,taking into account the specifics of the problems posed.For the software implementation of the proposed architecture,the Python programming language and the TensorFlow framework were utilized.The use of the proposed model is effective even in conditions of a limited amount of source data.Also as a result of experiments,a high degree of repeatability of the results was noted.According to the metrics,Mask R-CNN gave the high detection and segmentation results showing 0.9214,0.9876,0.9571 precision,recall,and F1-score respectively in road damage detection,and Intersection over Union(IoU)-0.3488 and Dice similarity coefficient-0.7381 in segmentation of road damages.

Modeling and analytical calculation of road damage coefficient considering tire pressure and damping of vehicles

The road damage coefficientβis a significant indicator to estimate the degree of the road damage caused by vehicles.The existing calculation method ofβis not convenient for the engineering application.To effectively evaluate the damage degree and facilitate the engineering application,this paper studied the simple and reliable analytical calculation method ofβ.Firstly,a dynamic model of the vehicle–road system was created.The tire pressure and the tire damping were considered in the model.Moreover,the relationship between the tire vertical stiffness and the tire pressure is approximated as a linear function.Secondly,based on the dynamic model,according to the definition ofβ,a concise analytical formula ofβwas derived and verified by numerical simulations.The relative errors of the analytical calculation results are all less than 0.1%.Thirdly,the influences of the tire pressure p,the damping ratioξs of the suspension system,and the damping ratioξt of the wheel system onβwere analyzed.Moreover,based on the analytical formula ofβ,a mathematical model of the optimal damping matching for the suspension system was established and a case study was also given.The research results show that the larger the tire pressure p,the larger the value ofβis.For each p,there is an optimal damping ratioξs.If the tire damping is ignored,it will lead to the design error forξs.Finally,some important conclusions were drawn.The analytical formula ofβand the conclusions can provide valuable references for the analysis of the road damage and the initial design of vehicle suspensions.

Comparison of deep convolutional neural network classifiers and the effect of scale encoding for automated pavement assessment

Deep learning has received a growing interest in recent years for detecting different types of pavement distresses and automating pavement condition assessment.A proper choice of deep learning models is key for successful pavement assessment applications.In this study,we first present a comprehensive experimental comparison of state-of-the-art image classification models to evaluate their performances on 11 pavement objects classification.Our experiments are conducted in different dimensions of comparison,including deep classifier architecture,effects of network depth,and computational costs.Five convolutional neural network(CNN)classifiers widely used in transportation applications,including VGG16,VGG19,ResNet50,DenseNet121,and a generic CNN(as the control model),are tested with a comprehensive pixel-level annotated dataset for 11 different distress and non-distress classes(UCF-PAVE 2017).In addition,we investigate a simple yet effective approach of encoding contextual information with multi-scale input tiles to classify highly random pavement objects in size,shape,intensity,texture,and direction.Our comparison results show that the multi-scale approach significantly improves the classification accuracy for all compared deep classifiers at a negligible extra computational cost.Finally,we provide recommendations of how to improve the classification performance of deep CNNs for automated pavement condition assessment based on the comparison results.