Cognitro Analytics is proud to have completed the testing and the development of a cloud-based traffic accidents predictive modeling tool, using the city of Abu Dhabi, UAE as a proof-of-concept to demonstrate its use and effectiveness in road traffic emergency management. In the first half of 2017, UAE road accidents claimed 315 lives and over 3,000 people suffered injuries, compared to 3,447 in the first half of 2016 only, according to the Khaleej Times article of July 30, 2017.
The predictive modeling development was based on RStudio, an open source machine learning library and Shiny, a GUI development toolkit. The tool is intended for usage by different authorities responsible emergency response, including the city municipality, to identify the accident hotspots and optimize response time by dispatching emergency vehicles to high-risk zones. It can be integrated over the current emergency services call center or similar applications. It can also be used by road and traffic authorities for city planning and resources allocation.
Leveraging advanced AI algorithms, we mapped past and predicted accidents on geospatial grids covering city roads and streets. The tool features powerful visualization capabilities that allow the user to view historical trends and patterns of accidents by severity across time and space as well as zooming on different areas of the city in order to maximize the knowledge gained from every incident reported on daily basis.
Descriptive analytics produces summary accidents statistics, which clearly shows correlation between accidents and patterns associated with seasonal events affecting traffic such as school holidays and major touristic events.
Nevertheless, it’s the predictive analytics component of the tool that makes it uniquely positioned to support strategic planning decisions. Utilizing a variety of data and measurements such as incident location and weather data, the tool allows the user to fast forward into a future window of time and predict the intensity and the location of potential accidents.
The tool does not only promise to bring significant cost savings and increased productivity to emergency services but will eventually help to save lives and support smart emergency services.
The development of this tool is part of Cognitro Analytics commitment to help create viable AI-driven and cost-effective solutions, harnessing the power of open source machine learning tools for safer roads and smarter cities.
For more questions or to see a demo, please contact us. firstname.lastname@example.org