TY - JOUR
T1 - Optimising Deep Learning at the Edge for Accurate Hourly Air Quality Prediction
AU - Wardana, I Nyoman Kusuma
AU - Gardner, Julian W.
AU - Fahmy, Suhaib Ahmed
N1 - KAUST Repository Item: Exported on 2021-02-10
Acknowledgements: This work was supported in part by Indonesia Endowment Fund for Education (LPDP), Ministry of Finance, Republic of Indonesia under grant number Ref: S-1027/LPDP.4/2019.
PY - 2021/2/4
Y1 - 2021/2/4
N2 - 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.
AB - 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.
UR - http://hdl.handle.net/10754/667290
UR - https://www.mdpi.com/1424-8220/21/4/1064
U2 - 10.3390/s21041064
DO - 10.3390/s21041064
M3 - Article
C2 - 33557203
SN - 1424-8220
VL - 21
SP - 1064
JO - Sensors
JF - Sensors
IS - 4
ER -