Structural Health Monitoring as a Service
Today creating data platforms for structural health monitoring is a complex, time consuming, and expensive task. In order to draw value from data gathering, the systems need to ingest large amounts of data, and make modern data analysis available at a large scale.
At University of Copenhagen I had the opportunity to write a thesis about how to create a data platform for structural health monitoring in the cloud. Instead of creating platforms for the individual bridge or structure, creating the platform as a service would allow the time spent on making the systems work, to be used for features, stability, and analysis.
I have in my thesis tried to solve the needs of modern Structural Health Monitoring Systems, by providing the following features:
- Gathering and storing data from structures.
- The data will generally be sensor data about the condition, motion, vibration, temperature, etc.
- Generating alarms when sensors breach thresholds set by maintenance personnel.
- Usually, buildings or structures will have predefined limits and design specifications that help to determine if the structure is structurally sound and that there are no current structural problems that require attention. These definitions will be converted into thresholds for alarms, that will alert the personnel with events that they should look more closely into.
- Showing a live view of the structures in customizable dashboards.
- Just as looking at a structure can help determine if there currently is a problem, getting a current view of sensor data in a user-friendly dashboard, or interactive model can help personnel to spot problems, or verify that everything is as it should be.
- Storing all data for later analysis.
- Data has value, and data analysis is only becoming more advanced and accessible. Storing the data will both let the personnel look at interesting historical events, and give the ability to do further data analysis in the future. Today getting value from the structural data can be a difficult task, but in the thesis, I suggest that the best way to solve this difficulty is by providing the best modern approaches to data analysis and exploration. These tools can then be used to automatically generate reports, or they can be handed over to data analysts to explore and analyze the data.
Based on my thesis I am starting a company called SenMoS making the product available to current and future structures. I am building the data platform in collaboration with Krabbenhøft & Ingolfsson, where it will be available as a part of their Structural Health Monitoring Solutions in the future.