A digital twin is a digital model of an intended or actual real-world physical product, system, or process that serves as the effectively indistinguishable digital counterpart of it for practical purposes, such as simulation, integration, testing, monitoring, and maintenance.
At Mountice Engineering we have developed several digitial twins by employing real-world data to teach an AI core to model the product, system, process.
Through the use of a single point interpolation algorithum and a fractal transformation the AI is capable of both decision making and proxy analysis even with a small set of training data.
Once twinned the performance of the product, system or process can be evaluated and tweaked at a fraction of the cost of traditional prototyping.
Real-life measurements can then be compared with expected values from the twin to evaluate the accuracy of the modeling.
When measured data is used as input it can be compared with expected performance against the original baseline.
Steven Ladouceur P.Eng. sladouceur_remove@mountice.com
The digital twin is system agnostic, it simply learns the behaviour of the product,system or process using the training data.
| Training Data | |||
|---|---|---|---|
| Input 1 | Input 2 | Output | Actions |
| 30 | 30 | 30 | Delete |
| 30 | 30 | 30 | Delete |
| 10 | 23 | 30 | Delete |
| 5 | 6 | 7 | Delete |
| 7 | 7 | 7 | Delete |
| 0 | 0 | 10 | Delete |
| 0 | 0 | 0 | Delete |
| 0 | 0 | 0 | Delete |
| 0 | 0 | 0 | Delete |
| 0 | 0 | 0 | Delete |
| 0 | 0 | 0 | Delete |
| 23 | 3 | 3 | Delete |
| 0 | 0 | 0 | Delete |
| 0 | 0 | 0 | Delete |
| 0 | 0 | 0 | Delete |