According to the World Health Organization, over 1.35 million people die in road crashes each year - making it one of the leading causes of death worldwide. Oftentimes, these accidents are caused by human error and can be prevented with better safety protocols. However, traditional methods such as manual analysis or physical testing can be costly and time-consuming.
With the United Nations Sustainable Development Goal (SDG) to reduce road fatalities and injuries, big data and Artificial Intelligence (AI) have become essential tools for improving road safety. By utilising innovative technology and data-driven decision-making, governments can better manage their roads and help to achieve SDG targets.
In this article, we’ll take a look at the ways that big data and AI are being used in the fight to improve global road safety. Learn how they are being used to inform policy decisions, identify risk areas, improve infrastructure design, and save lives.
In order to reduce the number of accidents on Australian roads, the government has been turning to artificial intelligence (AI) and big data for help. In the past, crash locations were treated after crashes had occurred. With the help of AI and Big Data, a more predictive approach is possible where potential crash locations are risk assessed and mitigated before lives are lost.
The AiRAP initiative was conceived by iRAP in 2019 to improve access to and application of existing and emerging data sources globally, including advances in artificial intelligence, machine learning, vision systems, LiDAR, telematics and other data sources. AiRAP refers to the ‘accelerated and intelligent’ capture of road safety-related data using automatic, repeatable and scalable methods to support road safety assessment, crash risk mapping, and investment prioritisation for all road users.
To demonstrate the effectiveness of AiRAP, iMOVE — an international research partnership — conducted a two-year project in 2020 that compared traditional methods for improving road safety against AiRAP technology. Results showed that AiRAP was more effective at improving data sharing amongst stakeholders and reducing the number of road accidents. The AiRAP technology was also found to cost significantly less than traditional methods.
The two types of source data used in the project (MN-R and MoMa) were demonstrated to be capable of capturing most AusRAP data attributes using either fully automated or accelerated (partially automated) methods. A total of seven attributes from TomTom’s MN-R data and 34 attributes extracted by Anditi from TomTom’s MoMa data were accredited. This means that the feature extraction techniques used to produce data for these road attributes have been confirmed to meet iRAP’s global standard and can be applied anywhere the source data is available.
So far, AiRAP has been shown to be effective in reducing accident rates by up to 50%. This is a significant improvement that could save many lives each year. This research and development project improved road infrastructure, data efficiency and accuracy and delivered substantial, long-term cost and efficiency savings to TfNSW.
Technology is playing an increasingly important role in UN global road safety goals. In particular, Artificial Intelligence (AI) and Big Data are being leveraged to enhance safety analysis.
For example, AI can predict traffic patterns and identify potential hazards. Additionally, Big Data can be used to track trends and help develop strategies to improve road safety.
The iMOVE project aimed to explore potential ‘off-the-shelf’ data sources, such as LiDAR and probe data, which could be used to accelerate road safety assessments from the manual coding of 50 safety attributes per 100m to an automated system.
The iMOVE project utilised Australian Big Data Analytics company, Anditi’s 3D portal RoadViewer, a world-first iRAP-accredited mobile LiDAR and imagery inspection system that opens the pathway to intelligent road safety analysis.
The iMOVE project pioneered positive implications for TfNSW, other Australian jurisdictions and internationally. The project aligned with the UN’s Second Decade of Action for Road Safety and coincided with the release of Australia’s National Road Safety Strategy and NSW’s Road Safety Plan. All three shared Star Rating targets as a critical road safety management and reporting component.
Data drives road risk ratings on an unprecedented scale. The project findings could pave the way for global road safety analysis indefinitely. The table below highlights progress made across a number of areas since the project commenced.
For the past four years, Anditi has been researching and developing automated and accelerated techniques to utilise mobile LiDAR and 360-degree imagery for the Star Rating of roads. A total of 52 attributes were able to be extracted by Anditi from TomTom’s MoMa data, with 34 of these being accredited under AiRAP
AI and other technologies used by Anditi's RoadViewer software to derive attributes for Star Rating of roads include LiDAR point cloud processing, image processing/computer vision algorithms, and fusion of information extracted from imagery and point cloud.
RoadViewer is the first to use these techniques to generate a star rating for a road network automatically. The software has been successfully tested in a Victoria, Australia, pilot study and leveraged for the iMOVE project. RoadViewer is built on top of the existing International Road Assessment Program (iRAP) protocol, which provides a comprehensive safety assessment of road infrastructure.
RoadViewer uses AI-powered image processing algorithms to automatically detect and measure assets such as Guard Railings, Curbs, Intersections and Median Islands.
This ground-breaking technology has made it possible to rapidly rate roads more accurately, with increased precision and consistency, without manual intervention - improving the safety of road users across Australia's highways.