Implementing a Volunteer Notification System into a Scalable, Analytical Realtime Data Processing Environment

Abstract

The pace at which next-generation Internet of Things networks, consisting of wirelessly distributed sensors and devices, are being developed is speeding up. More and more devices produce data in automated manners and the demand of smartphones and wearable devices is continuously increasing. With respect to volunteer notification systems (VNS), the resulting vast amounts of data can be utilized for profiling and predicting the whereabouts of people that, combined with machine learning algorithms, complement artificial intelligence (AI)-based decision systems. Hence, VNS benefit from keeping pace with the current developments by using the corresponding data streams in order to improve decision making during the volunteer selection process. In emergency scenarios, the velocity, low latency and reaction times of the system are essential, which results in the need of online stream-processing and real-time computational solutions. This paper will focus on a basic concept for implementing a VNS approach into a scalable, fault-tolerant environment that uses state-of-the-art analytical tools to process information streams in real-time as well as on demand, and applies machine learning algorithms for an AI-based volunteer selection. This work concentrates on leveraging open source Big Data technologies with the aim to deliver a robust, secure and highly available enterprise-class Big Data platform. Within the given context, this work will furthermore give an insight on state-of-the-art proprietary solutions for Big Data processing that are currently available.