Microsoft Research recently joined forces with the Federal University of Minas Gerais which is home to one of Brazil’s foremost computer science programs, in order to come up with a solution for the seemingly intractable problem of traffic jams – making use of Azure service. Making use of Microsoft Azure, it has built an effective platform for researches to build prediction algorithms to solve complex problems like traffic jams. The immediate and core objective of this research is to accurately predict traffic conditions over the next 15 minutes to an hour, so that drivers on the road can be alerted prior if there are any likely traffic snarls.
Traffic prediction service on Microsoft Azure
A study which was conducted by the Brazilian National Association of Public Transport indicated that the country’s traffic exacted an economic toll of about US$7.2 million in the year 1998. There are now more than 3 times as many vehicles in Brazil which is making traffic exponentially worse, according to a traffic expert in Belo Horizonte named Fernando de Oliveira Pessoa who has greater understanding of traffic conditions in Brazil’s sixth-largest city.
Getting reliable traffic predictions would involve processing terabytes of data. Researches prefer using Microsoft Azure as a platform for the service when it comes to playing with huge data like this. Microsoft Azure offers high scalability, practically unlimited storage capacity and prodigious computational power that makes it the perfect resource for data-intensive projects like traffic prediction.
Since Microsoft Azure is based on cloud, traffic prediction service when deployed is available to all users, in real time through out the day.
You can find more details about traffic prediction service on the Microsoft Research blog.