Enhancing Bulk KPI Evaluation Efficiency in Smart Cities Using Parallel Computing

The rapid proliferation of Internet of Things (IoT) devices, projected to reach over 16 billion in 2023, has significantly transformed sectors like Smart City management. This growth introduces challenges related to the volume, velocity, and variety of sensor data, necessitating efficient monitoring and management solutions. While existing tools and frameworks simplify data ingestion and storage, they often lack analytical capabilities, placing a greater burden on users for decision-making. Key Performance Indicators (KPIs) offer a promising approach to addressing this gap, enabling automated performance assessment across various dimensions such as sustainability and efficiency. This paper focuses on optimizing KPI evaluation routines within IoT management platforms, using the Logimic's smart city solution ACADA as a case study. The model and KPI evaluation code were first optimized to improve overall efficiency. This optimization alone resulted in a 13x speedup, reducing the evaluation time from 50 seconds to just 3.6 seconds. Subsequently, by leveraging multi-core architectures, performance was further enhanced. Utilizing four cores, a significant 27x speedup was achieved, ultimately bringing the evaluation time down to just 1.8 seconds. These results highlight the potential of such optimizations to enhance platform efficiency, reduce computational costs, and improve the scalability of IoT solutions for Smart Cities. However, it also emphasizes the importance of model and code quality, which is essential for efficient parallel computing.