Project Description
Professor Camata Visits Porto
Next week, Professor Guido Camata travels to Portugal! Prof. Camata will be at the University of Porto on March 21st to give a seminar. The research talk will present a new Digital Twinning technology that uses digital sensors, artificial intelligence algorithms, and finite element models.
The system creates a digital replica that is updated in real-time via data gathered by sensors in order to evaluate real-time health and predict the future response of a structure by simulating abnormal stress or damage conditions for maintenance, repair, and reinforcement.
The MonStr system of digital sensors (patent pending) allows over 2000 channels to be synchronized to less than 1ms. The system was developed to read up to 4kH of continuous data samples without the use of time windows. The MonStr nodes contain accelerometers, gyroscopes, triaxial magnetometers, biaxial inclinometers, temperature sensors, and a signal acquisition device.
In addition to internal sensors, the node can be connected to external sensors such as strain gauges, LVDTs, load cells, etc. The acquisition device not only reads the signal but is programmed to select and clean the data to be transmitted, greatly reducing the influx of data.
The MonStr system is fully integrated with the Opensource OpenSees computing program and artificial intelligence algorithms through the use of the STKO software. The artificial intelligence algorithms allow for Optimal Sensor Placement (OSP) and model calibration. They also interpret the damage and emit alarm signals that are then checked against the finite element model to assess whether the signal is a false positive. This reduces the number of false positives, especially during the first years of use when the data pool acquired is not sufficient for reliable interpretation using artificial intelligence.