Households are connected to better quality work, healthcare, and educational possibilities by road infrastructure. The World Bank has built or renovated almost 260,000 kilometers of roads since 2002, lending more money for roads than for social services, health care, and education put together. The World Bank uses satellite imagery from the Registry of Open Data on AWS, geospatial imaging, and Alteia data analytics enabled by Amazon Web Services (AWS) to evaluate road infrastructure more quickly and affordably. In nations like Mexico, Peru, and Tunisia, this aids local governments in repairing road infrastructure and establishing connections between marginalized populations.
The World Bank is the largest multilateral development bank in the world, with 189 member nations, and its goal is to improve shared prosperity and decrease poverty in a sustainable manner. It helps governments in developing nations raise the standard of living for their citizens and boost their economies by lending them money and offering technical assistance. By making economic and social opportunities more accessible, especially to historically underserved and inaccessible populations, transportation infrastructure plays a crucial role in achieving these goals.
The World Bank Group collaborated with Alteia, a top enterprise artificial intelligence (AI) software provider with its headquarters located in Toulouse, France, to develop an application on the Aether software platform. Based on publicly available data, the application, which runs on AWS, accurately harvests insights from road networks at scale in poor nations. Alteia effectively extracts road network data from satellite data using machine vision and geospatial images. Data streams from many sources are quickly aggregated, contextualized, and analyzed using the Aether platform. This makes it possible for Alteia to support use cases such as road network extraction, road network quality validation, and network data enrichment with other kinds of information.
It can be expensive and time-consuming to gather information about road networks in remote areas using high-resolution satellite imagery and conventional, on-the-ground road infrastructure surveys. The technology created by Alteia enables mapping and evaluating a nation’s whole road network for 15% less than optical inspection techniques, all thanks to the power of machine learning (ML) and geospatial data on AWS. As a result, 500,000 kilometers of roads may be mapped and evaluated in just three weeks as opposed to 15 years with conventional procedures. Alteia’s method was more than 90% less expensive than using high-resolution satellite imagery since it made use of AWS and multiple free data sources.
Assessing the dangers posed by outside events like flooding is also made easier by the computerized depiction of the road network. By adopting a risk-based decision-making process instead of a purely statistical one, governments can proactively handle possible issues and improve the resilience of their infrastructure. By utilizing satellite imagery data from the Registry of Open Data on AWS, data preparation, and machine learning algorithms running on Amazon Elastic Kubernetes Service (Amazon EKS), Alteia was able to successfully enable access to enormous volumes of satellite images for nations like Mexico, Peru, and Tunisia. For every nation, the program produced up to 2 terabytes (TB) of refined data, which Amazon Simple Storage Service (Amazon S3) effectively stored and distributed.
Assessing road infrastructure at scale
The Alteia program cross-references vector road data, such as OpenStreetMap (OSM), with Sentinel imagery from the Registry of Open Data on AWS. Alteia builds an improved road network with extra features and contextual attribution, such demographic data, using AWS computation and infrastructure. To assess road roughness and distinguish between paved and unpaved roads, Alteia employs optical imagery from Sentinel 2 and Synthetic Aperture Radar (SAR) satellite photography. Crowdsourced ground truth data is used to verify the accuracy of the data.
Yann Ameho, Alteia’s chief technology officer, stated, “We always aim to get the best performance for our customers, and we can do that by leveraging AWS EKS to deploy our algorithms at scale.”
Alteia used Amazon EKS, a managed Kubernetes cluster solution, to create a microservices architecture. In order to store massive geospatial datasets and facilitate faster analysis through effective data access, this application effortlessly combines data from third-party sources with Amazon S3. Alteia uses Amazon CloudWatch for monitoring and logging and Amazon GuardDuty for threat detection to ensure the security and observability of their platform.
The Alteia application deploys several container pods tailored for particular tasks using Amazon EKS running on Amazon Elastic Compute Cloud (Amazon EC2) servers. Alteia developed a microservice for producing tiles using vector data and satellite imagery. For speedy retrieval and scalability, vector tiles were kept in a MongoDB database while raster tiles were cached on Amazon S3.
Raster and vector tile servers were integrated into the application to supply data for the analytics and machine learning components, which are utilized for data enrichment and inference. End users had a smooth user experience thanks to the web server’s real-time visualization capabilities.
Governments can prioritize maintenance tasks, maximize spending, and improve their roadways by efficiently calculating and evaluating road networks. By facilitating connections between marginalized populations, this program significantly advances accessibility and transportation infrastructure. Aether also makes it possible to prioritize road investments in lending operations, create robust networks, and track the development of infrastructure between nations. Decision-makers can now do statistical analysis on how infrastructure affects socioeconomic outcomes thanks to this geospatial solution.
The assessment of extensive road networks has been completely transformed by the convergence of AI, cloud-based data availability, and robust cloud computing. The creative application of geospatial analytics, made possible by the World Bank’s purpose, Alteia’s experience, and AWS cloud capabilities, has enabled governments and businesses to make data-driven decisions and build more robust and effective road networks across the globe.
