The tire–road friction coefficient is a critical evaluation index of the service performance of roads: it governs the stopping distance, traction control, and stability of vehicles. Moreover, friction information is also needed in many function units of modern vehicles. This paper proposes a novel data-driven approach for inference of the maximum tire–road friction coefficient using a combination of vehicle dynamics signal and machine vision data. The approach is aimed at robust road condition perception that can provide frequent measurements over large areas across all weather conditions. Two different neural network architectures were adopted to extract in-depth features behind vehicle dynamics signals and road surface images. Features from these two types of data were then fused in two different levels, namely feature level and decision level, forming a multi-feature fusion neural network. The proposed network performs better than models based only on dynamic signals or vision data. The method proposed was applied to real data obtained from an electric car in a highway driving scenario. For classification of the maximum tire–road friction coefficient, the proposed network can yield F1-score increments of 0.09 and 0.18 from dynamics-based and vision-based sub-models, respectively. For the maximum tire–road friction coefficient value regression, the proposed model also achieves the highest R-square score of 0.71. Of these two types of data collected under highway driving scenarios, the vision data contribute more to the overall performance of the proposed model. Nevertheless, the dynamics data possess excellent potential in poor lighting conditions. With these fused features, the proposed multi-feature fusion network can not only improve the accuracy of maximum tire–road friction coefficient estimation but also is deemed workable for a broader range of scenarios.

Modern cars are equipped with many sensors that measure information about the vehicle and its surroundings. These measurements are therefore related to the ride-surface conditions over which the vehicle is passing. The paper commences by outlining a four-component vision for performing road condition evaluation based on in-vehicle sensor readings and subsequent feeding of pavement management systems (PMSs) with live condition information. This is expected to enrich the functionalities of PMSs, and ultimately lead to improved maintenance and repair decisions. Next the LiRA (Live Road Assessment) project—an ongoing proof-of-concept attempt to realize the vision components—is presented. The project elements and software architecture are described in detail, listing any assumptions, compromises, and challenges. LiRA involves a fleet of 400 electric cars operating in Copenhagen, both within the city streets and nearby highways. Sensor data collection is performed with a customized Internet of Things (IoT) device installed in the cars. Data processing and structuring involve new software tools for quality control, spatio-temporal interpolation, and map matching. A hybrid approach, combining machine learning models with physical (mechanics-based) models, is currently being applied to convert data into relevant information for PMSs. Based on the experience thus far with LiRA, the vision actualization is deemed achievable, workable, and up-scalable.

The road transportation sector is a dominant and growing energy consumer. Although investigations to quantify the road infrastructure’s impact on energy consumption have been carried out, there are currently no standard methods to measure or label the energy efficiency of road networks. Consequently, road agencies and operators are limited to restricted types of data when managing the road network. Moreover, initiatives meant to reduce energy consumption cannot be measured and quantified. This work is, therefore, motivated by the desire to provide road agencies with a road energy efficiency monitoring concept that can provide frequent measurements over large areas across all weather conditions. The proposed system is based on measurements from in-vehicle sensors. The measurements are collected onboard with an Internet-of-Things (IoT) device, then transmitted periodically before being processed, normalized, and saved in a database. The normalization procedure involves modeling the vehicle’s primary driving resistances in the driving direction. It is hypothesized that the energy remaining after normalization holds information about wind conditions, vehicle-related inefficiencies, and the physical condition of the road. The new method was first validated utilizing a limited dataset of vehicles driving at a constant speed on a short highway section. Next, the method was applied to data obtained from ten nominally identical electric cars driven over highways and urban roads. The normalized energy was compared with road roughness measurements collected by a standard road profilometer. The average measured energy consumption was 1.55 Wh per 10 m. The average normalized energy consumption was 0.13 and 0.37 Wh per 10 m for highways and urban roads, respectively. A correlation analysis showed that normalized energy consumption was positively correlated to road roughness. The average Pearson correlation coefficient was 0.88 for aggregated data and 0.32 and 0.39 for 1000-m road sections on highways and urban roads, respectively. An increase in IRI of 1 m/km resulted in a 3.4% increase in normalized energy consumption. The results show that the normalized energy holds information about the road roughness. Thus, considering the emergence of connected vehicle technologies, the method seems promising and can potentially be used as a platform for future large-scale road energy efficiency monitoring.