Patterns for real-world-aware and real-time solutions
IBM Systems Journal
| Hi-index | 0.00 |
This paper describes an IBM and the University of Miami joint research project providing infrastructure and application components to advance structural health monitoring of civil infrastructures such as bridges. We discuss a newly developed software infrastructure that enables a shift from batch to continuous bridge monitoring, providing continuous real-time sensor data collection and forwarding to a monitoring location where the data is cleansed (e.g., corrected), normalized, and recorded. We also discuss how significant load events are detected and fracture mechanics analytics are applied to assess bridge structural health. Finally, we discuss visualization of the raw and processed data. The University of Miami civil engineers developed the continual analytics techniques based on fracture mechanics and acoustic-emission analytics that together with the software infrastructure make it possible to perform more accurate and real-time monitoring of bridge deterioration. For civil structural health monitoring, this introduces a shift from time- to event-based analytics by exploiting a rolling, potentially complex, significant event-based threshold that triggers the analytics. The system was deployed on a single laboratory test specimen to assess the validity of the messaging and analytical tools in a simulated bridge environment.