A Context-Sensitive Cloud-Based Data Analytic Mobile Alert and Optimal Route Discovery System for Rural and Urban ITS Penetration

The rapid growth in the number of road users and poor road management have been deemed responsible for the upsurge in road congestions and fatalities in recent times. Many of the lives lost was due to inadequate or inefficient public-accessible alerts system and rerouting mechanisms during emergencies. The Intelligent Transportation System (ITS) was anticipated as a solution to the numerous road networks usage problems. Recently, some developed countries have implemented some forms of ITS initiatives. But the transition of the road networks to a fully integrated ITS has been slow and daunting due to the huge cost of implementation. The use of mobile devices as backbone infrastructure for ITS networks during public emergencies has been proposed. Despite the advantage of being a cheap alternative, low computing power of mobile devices limit their potentials to support the expected Big Data ITS traffic. In this paper, we propose a cloud-based context-sensitive ITS infrastructure that uses the cloud as a primary aggregator of traffic messages plus a hybrid Data Analytics algorithm. The algorithm combines the enhanced features of Apache-Spark and Kafka frameworks blended with collaborative filtering using the ensemble machine learning classifier. The novelty of our approach stems from its ability to provide load balancing routing services based on the users’ profiles, and avoid congestion-using the Dynamic Round Robin scheduling algorithm to reroute users with similar profiles.