Next Article in Journal
Using a Low-Power Spiking Continuous Time Neuron (SCTN) for Sound Signal Processing
Next Article in Special Issue
EDISON: An Edge-Native Method and Architecture for Distributed Interpolation
Previous Article in Journal
A Privacy-Preserving Key Management Scheme with Support for Sybil Attack Detection in VANETs
Previous Article in Special Issue
Joint Content Placement and Storage Allocation Based on Federated Learning in F-RANs
Article

Optimising Deep Learning at the Edge for Accurate Hourly Air Quality Prediction

1
School of Engineering, University of Warwick, Coventry CV4 7AL, UK
2
Computer, Electrical and Mathematical Sciences and Engineering, King Abdullah University of Science and Technology (KAUST), Thuwal 23955, Saudi Arabia
3
Department of Electrical Engineering, Politeknik Negeri Bali, Badung, Bali 80364, Indonesia
*
Author to whom correspondence should be addressed.
Academic Editors: Claudio Savaglio, Salvatore Venticinque, Atay Ozgovde and Teemu Leppänen
Sensors 2021, 21(4), 1064; https://doi.org/10.3390/s21041064
Received: 24 December 2020 / Revised: 28 January 2021 / Accepted: 29 January 2021 / Published: 4 February 2021
(This article belongs to the Special Issue Sensors and Smart Devices at the Edge: IoT Meets Edge Computing)
Accurate air quality monitoring requires processing of multi-dimensional, multi-location sensor data, which has previously been considered in centralised machine learning models. These are often unsuitable for resource-constrained edge devices. In this article, we address this challenge by: (1) designing a novel hybrid deep learning model for hourly PM2.5 pollutant prediction; (2) optimising the obtained model for edge devices; and (3) examining model performance running on the edge devices in terms of both accuracy and latency. The hybrid deep learning model in this work comprises a 1D Convolutional Neural Network (CNN) and a Long Short-Term Memory (LSTM) to predict hourly PM2.5 concentration. The results show that our proposed model outperforms other deep learning models, evaluated by calculating RMSE and MAE errors. The proposed model was optimised for edge devices, the Raspberry Pi 3 Model B+ (RPi3B+) and Raspberry Pi 4 Model B (RPi4B). This optimised model reduced file size to a quarter of the original, with further size reduction achieved by implementing different post-training quantisation. In total, 8272 hourly samples were continuously fed to the edge device, with the RPi4B executing the model twice as fast as the RPi3B+ in all quantisation modes. Full-integer quantisation produced the lowest execution time, with latencies of 2.19 s and 4.73 s for RPi4B and RPi3B+, respectively. View Full-Text
Keywords: air quality prediction; PM2.5 prediction; deep learning; post-training quantisation; edge computing air quality prediction; PM2.5 prediction; deep learning; post-training quantisation; edge computing
Show Figures

Graphical abstract

MDPI and ACS Style

Wardana, I.N.K.; Gardner, J.W.; Fahmy, S.A. Optimising Deep Learning at the Edge for Accurate Hourly Air Quality Prediction. Sensors 2021, 21, 1064. https://doi.org/10.3390/s21041064

AMA Style

Wardana INK, Gardner JW, Fahmy SA. Optimising Deep Learning at the Edge for Accurate Hourly Air Quality Prediction. Sensors. 2021; 21(4):1064. https://doi.org/10.3390/s21041064

Chicago/Turabian Style

Wardana, I N.K., Julian W. Gardner, and Suhaib A. Fahmy. 2021. "Optimising Deep Learning at the Edge for Accurate Hourly Air Quality Prediction" Sensors 21, no. 4: 1064. https://doi.org/10.3390/s21041064

Find Other Styles
Note that from the first issue of 2016, MDPI journals use article numbers instead of page numbers. See further details here.

Article Access Map by Country/Region

1
Back to TopTop