Result Details
Enhancing Bulk KPI Evaluation Efficiency in Smart Cities Using Parallel Computing
John Petr, Ing., DIFS (FIT), DCGM (FIT)
Hynek Jiří, Ing., Ph.D., DIFS (FIT)
Valný Michal
Hruška Tomáš, prof. Ing., CSc., DIFS (FIT)
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.
Key Performance Indicators, Smart Cities, FaaS, AWS, cost efficiency
@inproceedings{BUT197685,
author="Ondřej {Šulc} and Petr {John} and Jiří {Hynek} and {} and Tomáš {Hruška}",
title="Enhancing Bulk KPI Evaluation Efficiency in Smart Cities Using Parallel Computing",
booktitle="IEEE Xplore",
year="2025",
series="2025 Smart City Symposium Prague (SCSP)",
pages="1--6",
publisher="Institute of Electrical and Electronics Engineers",
address="Prague",
doi="10.1109/SCSP65598.2025.11037697",
isbn="979-8-3315-2550-7",
url="https://ieeexplore.ieee.org/document/11037697"
}