Dalarna University's logo and link to the university's website

Gravel roads connect sparse populations and provide highways for agriculture and the transport of forest goods. Gravel roads are an economical choice where traffic volume is low. In Sweden, 21% of all public roads are state-owned gravel roads, covering over 20,200 km. In addition, there are some 74,000 km of gravel roads and 210,000 km of forest roads that are owned by the private sector. The Swedish Transport Administration (Trafikverket) rates the condition of gravel roads according to the severity of irregularities (e.g. corrugations and potholes), dust, loose gravel, and gravel cross-sections. This assessment is carried out during the summertime when roads are free of snow. One of the essential parameters for gravel road assessment is loose gravel. Loose gravel can cause a tire to slip, leading to a loss of driver control.  Assessment of gravel roads is carried out subjectively by taking images of road sections and adding some textual notes. A cost-effective, intelligent, and objective method for road assessment is lacking. Expensive methods, such as laser profiler trucks, are available and can offer road profiling with high accuracy. These methods are not applied to gravel roads, however, because of the need to maintain cost-efficiency. 

In this thesis, we explored the idea that, in addition to machine vision, we could also use machine hearing to classify the condition of gravel roads in relation to loose gravel. Several suitable classical supervised learning and convolutional neural networks (CNN) were tested. When people drive on gravel roads, they can make sense of the road condition by listening to the gravel hitting the bottom of the car. The more we hear gravel hitting the bottom of the car, the more we can sense that there is a lot of loose gravel and, therefore, the road might be in a bad condition. Based on this idea, we hypothesized that machines could also undertake such a classification when trained with labeled sound data. Machines can identify gravel and non-gravel sounds. In this thesis, we used traditional machine learning algorithms, such as support vector machines (SVM), decision trees, and ensemble classification methods. We also explored CNN for classifying spectrograms of audio sounds and images in gravel roads. Both supervised learning and CNN were used, and results were compared for this study. In classical algorithms, when compared with other classifiers, ensemble bagged tree (EBT)-based classifiers performed best for classifying gravel and non-gravel sounds. EBT performance is also useful in reducing the misclassification of non-gravel sounds. The use of CNN also showed a 97.91% accuracy rate. Using CNN makes the classification process more intuitive because the network architecture takes responsibility for selecting the relevant training features. Furthermore, the classification results can be visualized on road maps, which can help road monitoring agencies assess road conditions and schedule maintenance activities for a particular road.

A well-maintained road network is essential for sustainable economic development, providing vital transportation routes for goods and services while connecting communities. Sweden's public road network includes a significant portion of gravel roads, particularly cost-effective for less populated areas with lower traffic volumes. However, gravel roads deteriorate quickly, leading to accidents, environmental pollution, and vehicle tire wear when not adequately maintained. The Swedish Road Administration Authority (Trafikverket) assesses gravel road conditions using subjective methods, analysing images taken during snow-free periods. Due to cost constraints, this labour-intensive process is prone to errors and lacks advanced techniques like road profilometers.

This thesis explores the field of assessing gravel road conditions. It commences with a comprehensive review of manual gravel road assessment methods employed globally and existing data-driven smart methods. Subsequently, it harnesses machine hearing and machine vision techniques, primarily focusing on enhancing road condition classification by integrating sound and image data.

The research examines sound data collected from gravel roads, exploring machine learning algorithms for loose gravel conditions classification with potential road maintenance and monitoring implications. Another crucial aspect involves applying machine vision to categorise image data from gravel roads. The study introduces an innovative approach using publicly available resources like Google Street View for image data collection, demonstrating machine vision's adaptability in assessing road conditions.

The research also compares machine learning methods with manual human classification, specifically regarding sound data. Automated approaches consistently outperform manual methods, providing more reliable results. Furthermore, the thesis investigates combining audio and image data to classify road conditions, particularly loose gravel scenarios. Early feature fusion using pre-trained models significantly improves classifier accuracy.

The research proposes using cost-effective devices like mobile phones with AI applications attached to car windshields to collect audio and visual data on gravel road conditions. This approach can provide more accurate and efficient data collection, resulting in real-time mapping of road conditions over considerable distances. Such information can benefit drivers, travellers, and road maintenance agencies by identifying problematic areas with loose gravel, enabling targeted and efficient maintenance efforts, and minimising disruptions to traffic flow during maintenance operations.