Vulnerability of critical road infrastructure increasingly gains awareness in transportation planning research. Natural hazards, such as earthquakes and floods, threaten road infrastructure and thus, the society that depends on these infrastructures. In case of a natural disaster one of the main challenges is to analyze as soon as possible the functionality of the road infrastructure and identify critical links within the road network.

Within the IPF-KRITIS project, a generic, multi-scale concept to analyze the vulnerability of critical road infrastructure is developed. This concept follows a modular approach: its basic module evaluates the accessibility of emergency. Further indices like for example the Accessibility and Remoteness Index of Australia (Taylor & Susilawati 2012) can be calculated. Additional modules provide the possibility to calculate the index based on a grid and to generate a degraded network scenario. The generic concept is applied exemplarily on two wildfire scenarios in Chile and in Portugal.

Figure 1: Emergency Facility Accessibility Index (EFAI) in the BioBio and Maule region in Chile before and after the 2017 wildfires and impact of the 2017 wildfires (Guth et al. 2019).

OpenStreetMap (OSM) is used as a worldwide and readily available data source. OSM is a map of the world which is collected and updated permanently by volunteers worldwide. Due to the crowdsourced nature of OSM however, the quality of the road network in OSM varies widely. Information required for routing like speed information are often missing and errors can be found for example due to misclassifications of roads. Misclassifications can lead to false assumptions about capacity, maximum speed or road quality which can hinder routing applications.

In the IPF-KRITIS project, methods are developed that aim at improving the navigability of OSM road data for the application in the generic concept. On the one hand, a module is developed which estimates average speed based on the OSM parameters road class, road surface, road slope and link length. Average speed vales are estimated based on a Fuzzy Control System which is applied exemplarily for two study regions in Chile and Australia. On the other hand, an additional module is developed that identifies road misclassification errors in OSM. It searches for gaps in hierarchical subnetworks which may indicate classification errors. Detours are used to estimate the error probability. This module helps to identify misclassifications which might cause large detours in routing applications.

Figure 2: Map of the difference between calculated speed and reference speed in the BioBio and Maule region in Chile and in the north of New South Wales in Australia (Guth et al. 2020).

The contributions of the IPF-KRITIS Project are:

• Development and implementation of a generic concept to analyze the vulnerability of critical road infrastructure (Guth et al. 2019);
• Development and implementation of accessibility indices (e.g. EFAI, ARIA) within the generic concept (Braun et al. 2018, Guth et al. 2019);
• Analysis of an evacuation process in Chile as a combination of social science and GIS methods (Kubisch et al. 2019, Kubisch et al. 2020);
• Development and implementation of a Fuzzy Control System to estimate average speed from multiple OSM parameters (Stötzer et al. 2019, Guth et al. 2020);
• Development and implementation of an approach to identify road misclassification errors in OSM.

The generic concept can be applied to quickly identify critical links in the road network after a natural disaster. The application of OSM data enables a quick analysis which is independent of local data availability of the affected countries. The software developed in the KRITIS project is published on GitHub under a creative commons license.

Johanna Guth (former Stötzer) is doing her PhD within the KRITIS project.