Resolving Traffic Congestion Problems With Big Data And Data Science
I'm going to use this blog series as an opportunity to share some of the knowledge I gained while creating an ITS for the city of Ho Chi Minh City (HCMC). In the coming blogs, I'll discuss some of the statistical methods employed and how "big data" was used to tackle this difficult issue; many of these methods are also useful for tackling business issues that arise in the real world.
As many of you may be aware, HCMC is one of SE Asia's largest and busiest cities, with one of the region's most intricate traffic systems and patterns, including a high density of motorcycles. In addition, several of the city's main thoroughfares lack lanes, creating congestion hotspots that only worsen during peak rush hours.
As a result, it is increasingly difficult to acquire traffic data due to its complexity. Therefore, an Intelligent Transportation System (ITS) is a promising remedy for handling such complexity. ITSs have been created and deployed in industrialized nations under various formations. However, they are rarely employed in most developing nations, primarily because they are expensive to build, administer, and maintain.
Check out the best data science course available online to familiarize yourself with big data tools.
At the John von Neumann Institute, we advocated the creation of a useful ITS that uses GPS data gathered from many types of transportation and is appropriate for developing nations like Vietnam. The city's Ministry of Transportation financed this project.
The ITS project has three primary business/operational goals:
The primary data sources for the ITS system are end-user smartphone GPS data and GPS data from onboard GPS devices, traffic cameras, traffic sensors, etc. As a result, unlike previous ITSs, this approach does not necessitate installing a system to gather data. This improved the ease of deployment for our system. But this also implied that we had to create Standardization criteria to transform the GPS data collected into a single class for further analysis.
The total solution has to be affordable for development and ongoing maintenance, much like any large-scale system. In this instance, we found that the operator's (the Ministry of Transportation) costs for creating, installing, and maintaining our system were significantly cheaper than those of current ITS models. We also concluded that the solution had to be simple for end customers to utilize. Fortunately, the cost of purchasing GPS devices for motorcycles in nations like Vietnam is inexpensive, and adoption of such devices is generally common, making them an extremely accessible source of data. Additionally, smart technologies (such as smartphones, tablets, etc.) are becoming more widely used and inexpensive in HCMC nowadays. Lastly, the price of creating mobile applications (to collect GPS data) is growing increasingly less expensive.
According to experimental findings, a 2–3% GPS device penetration rate among all drivers is sufficient to estimate traffic flow velocity on specific road segments reliably and effectively. A disproportionately large number of narrow roads are solely appropriate for motorcycles in developing nations like Vietnam. Therefore, it was deemed that our ITS could adequately cover most of the urban road networks in HCMC with the help of GPS motorbike data and GPS data from mobile applications.
Hope this article helped you in understanding how data science and big data help in resolving traffic problems in developing countries. As you can see, data is found almost everywhere. If you want to learn more about cutting-edge technologies, there are many best data science courses in India, where you can be certified by IBM.