Civil infrastructure systems are subject to aging with time, e.g. bridges. Such systems deteriorate performance wise and physically get damaged; the two translate into reduced service life and increased safety concerns of the infrastructure. This has been noted with due concern over the years and various monitoring and damage detection measures have been put in place. Narrowing down to bridges, bridge health monitoring has become a hotbed of inquiry owing to the high cost of construction, thus proper maintenance being the best alternative to lengthen the service life of such structures. As such, vibration-based health monitoring techniques have achieved promising progress thanks to the breakthrough in data analysis techniques and rapid development of wireless sensor networks.
Nonetheless, such wireless techniques have limitations that inhibit their wide scale application. With the current craze of smartphones and widespread ease of internet access, the idea of using smartphones to collect bridge health information has attracted increasing interest. Smartphones are often equipped with various sensors whose data can be uploaded in real-time to clouds from where it can be cloud-sourced for analysis. Unfortunately, to date, no study has been published on the notion of developing a framework for using a large number of smartphones on vehicles for bridge health monitoring and presenting analytical and experimental verifications.
In a recent publication, scientists at University of Alberta: Qipei Mei (PhD candidate) and Prof. Mustafa Gül from the Department of Civil and Environmental Engineering employed a novel crowdsourcing-based framework for indirect bridge health monitoring by utilizing data from smartphones in a large number of vehicles. In the proposed approach, smartphones were used as sensors from which the data collected was used to detect damage in bridges. Their work is currently published in the research journal, Structural Health Monitoring.
In brief, the proposed methodology was developed based on the Mel-frequency cepstral coefficients and Kullback–Leibler (KL) divergence. Data collected regarding specific bridge parameters, from the developed model, was subjected to numerical analysis in a bid to verify the proposed approach. Finally, a smartphone app developed for data collection was introduced and used in the laboratory experiments.
The authors showed that smartphones could potentially be used to indirectly detect damage in bridges. On technical terms, the proposed DFs calculated based on crowdsourcing data from smartphones in vehicles successfully identified the existence of damage. In addition, the research pair was also able to extract useful information about severity of the damages.
In summary, Qipei Mei-Mustafa Gül presented an interesting and novel framework for damage detection of bridges using crowdsourced data collected from a large number of vehicles passing through a bridge at different times. For purposes of the study, the team simulated the crowed-sourced vehicles by altering the configurations of a single vehicle. All in all, it was demonstrated that smartphones could be utilized for crowdsourcing based monitoring of bridges in the smart cities of future.
Qipei Mei and Mustafa Gül (2018). A crowdsourcing-based methodology using smartphones for bridge health monitoring. Structural Health Monitoring, page 1–18.Go To Structural Health Monitoring