Mobile Edge Intelligence for Enhancing Performance of Vehicular Air Quality Sensors

The objective of this master thesis project is to exploit Mobile Edge Computing Intelligence to enhance the performance of data processing for Vehicular Air Quality Sensors

Mobile sensors  which are widely deployed on buses, cars and bikes are getting attention for Air Quality Monitoring because measurements while moving can serve as input data to map the dynamics of air pollution: searching for hot spots of emissions, evaluating the exposure of air pollutants on individual level, or improving the temporal and spatial resolution of monitoring information. The data from mobile measurements can be utilized in the best way; for example: to create near real time air pollution maps with high spatial resolution that can help the citizens to choose the proper routes to travel, residential areas to buy apartments, etc…

Mobile sensors is based on a mobile platform mounted on Vehicles for data collecting, processing, modelling and transmitting wirelessly. There are significant problems of the existing mobile air quality sensors where almost raw data is transmitted through Vehicle-Cloud communication:

(1) Specific issues related to data collected from mobile monitoring devices, in-motion data collection. 

(2) Specific issues related to data communication between vehicles and cloud/server. 

(3) Specific issues related to calibration models to improve the accuracy for the AQ sensors: Data quality of the sensors is often questionable due to its geographical dependency; existing AQ sensors must be manually calibrated for a given deployment site, making them unsuitable for mobile deployment.     

As the wireless transmission of large data from mobile devices cannot be available, there is a need to process data locally. Edge computing [1] recently became a complementary approach for cloud computing by shifting some processing capacity to the edge network, hence reducing data traffic, cloud storage and computation. Just recently, edge processing has started to be used in mobile air monitoring systems [2, 3]. However, there will be additional issues related to Edge computing:

(4) specific issues related to data storage and processing capacities on mobile platforms. 

The objective of this master topic is therefore to exploit Mobile Edge Intelligence to enhance performance of Vehicular AQ Sensors in terms of data processing at the mobile edge to overcome the listed issues (1), (2), (3), and (4), some examples:

  1. Develop/Exploit the data processing framework that can be deployed on mobile sensors (low latency processing in limited  data storage and processing capacities) to manage real-time data fusion and analysis at the Edge.
  2. Develop/Exploit the collective edge intelligence based on aggregate information exchanged between cognitive neighboring sensors, urban features, climate conditions, etc.. (that effect on air pollution), which will be used for self-calibrated sensors (to improve monitoring accuracy) for a new deployment location, enabling mobile deployment.

The solutions will be tested in the mobile sensors network developed under the HAPADS project (an EEA Norway-Poland grant run by NILU)

Description of tasks

  • Literature review
  • Explore applicability of promising Edge intelligence for  data processing 
  • Device solution based on chosen Edge intelligence method(s)
  • Comparative evaluation 

Prerequisite

  • Distributed systems 
  • Solid coding skills

Contact information

  • send email to frank@ifi.uio.no and amirhost@ifi.uio.no

 

 

References

  1. N. Abbas, Y. Zhang, A. Taherkordi, T. Skeie, Mobile Edge Computing: A Survey, IEEE Internet of Things Journal, 5(1) (2018)
  2. Idrees, Zeba & Zou, Zhuo & Zheng, Lirong. (2018). Edge Computing Based IoT Architecture for Low Cost Air Pollution Monitoring Systems: A Comprehensive System Analysis, Design Considerations & Development. Sensors. 18. 3021. 10.3390/s18093021. 
  3. Moon J, Kum S, Lee S. A Heterogeneous IoT Data Analysis Framework with Collaboration of Edge-Cloud Computing: Focusing on Indoor PM10 and PM2.5 Status Prediction. Sensors (Basel). 2019;19(14):3038. Published 2019 Jul 10. doi:10.3390/s19143038

 

Publisert 12. okt. 2021 02:47 - Sist endret 12. okt. 2021 08:42

Omfang (studiepoeng)

60