Scientific paper ID 2075 : 2021/2

The advances of Information Technology (IT) in the 21st century have caused major and revolutionary changes in all sectors of human activities. IT has stopped being a factor only in computer systems and networks or in mobile communications. Digital technologies have penetrated everywhere with their capabilities to create and transmit data from any object. This model and technology are known nowadays as the Internet of Things (IoT) - connected heterogeneous objects on the Internet. Data sources from this paradigm can also be vehicles and transportation systems that report and monitor activities in the transport sector. Digital data created a world where big volumes of diverse data are being generated at great velocity. This created another phenomenon – Big data. The importance of Big data was quickly realized as it allows better and innovative type of analysis and optimization. Collected and processed data helps improving business operations and increases the efficiency all kind of practices. Such activities have been closely related to the ubiquitous data creation, which in turn gave rise to systems, capable of extracting, saving, processing, visualizing and analyzing Big data. This work presents one such data processing system that can be applied in many fields, including transportation. The system is constructed from blocks and modules that are scalable, open source, and integrate easily with the most popular and best well-known software products from the Hadoop environment. The presented modular system has been tested and its results demonstrate the applicability for transport data, taken from traffic events in the Los Angeles area. The data has been collected over a period of three months and consists of tens of thousands events. The system also allows for the collection, analysis, and visualization of other types of big data in the transport sector.