IEEE Future Directions
Related Articles
I had a call last week with people from Mevea, a Finnish company that is taking the concept of Digital Twin as the springboard for their business.
As a growing number of players in the industry, they see a Digital Twin as an integral part of a product, spanning its whole life cycle (see clip) and provide their software and services to company to transform their business into the Digital Space leveraging on Digital Twins.
They see a Digital Twin as a stand-alone life-like product that can be used for simulation purposes starting with the design phase and can also be used to discuss with the client how the product will look like and will operate -it is not a mock up, it has functional capability. A client might sit at a crane simulation of the future product and operate it seeing from himself if the operation would meet his needs.
They also see the Digital Twin as on-board with the product, keeping track of the operation parameters and possibly interacting with the physical twin (they are looking at integration of AI with the Digital Twin).
Furthermore, they see the Digital Twin as the on line copy of the product, kept in synch through the operation data received by the product. The same user interface that operates the physical twin is also generating operation command to the Digital Twin. The Digital Twin operates in the cyberspace through simulation of the physical twin based on the emulation of its physical characteristics. Note that the operation in the cyberspace takes place in a virtually replicated environment to be au pair with the conditions encountered by the physical twin.
The care taken to model the real operation environment is impressive. Mevea uses aerial photos (taken from Google maps and from drones filming) to replicate the operation ambient in the cyberspace.
As mentioned, they are looking at integrating AI with Digital Twins, expecting on the one hand the Digital Twin to generate data usable to train the AI component that in turns can be used for testing alternatives at the design phase as well as in planning an operation task. Machine learning can be used to suggest operation procedures and simulate the potential outcome. Additionally, data received from the physical twin can be analysed -and augmented- through AI.
They claim that using Digital Twins through the whole product life cycle is showing a 50% reduction in prototyping time, a 30% reduction in lead time, a 90% reduction in the control software production time and an increase of 25% in operation. These figures are derived from recent products development and operation.
