Last week, members of the SHAPE Architecture team attended the Digital Transformation / Digital Twin Summit hosted by BCIT. Numerous speakers from across industries shared their insights on how different initiatives could be taken with Digital Twin technologies in the future.
We found that Digital twins span many industries – from industrial processes (e.g. steel factories), to buildings (individual buildings to group clusters), all the way to civic infrastructure (e.g transit lines like the Canada Line). They offer immense value in testing various scenarios with little incremental cost compared to real-world simulations.
Our team highlighted a few learnings from the event.
- Digital twins offer safety and cost benefits for training compared to real-world training (the aviation industry has been using flight simulators similar to this for well over a decade)
- Digital twins offer immense value in testing various scenarios with little incremental cost compared to real-world simulations.
- Digital twins span many industries, from industrial processes (e.g. steel factories) to buildings (individual buildings like airports to groups of buildings such as post-secondary campuses or even small cities) to civic infrastructure (e.g transit lines like the Canada Line)
- Virtual Reality (VR) can be used to experience digital twin models.
- Augmented Reality (AR) is combining the digital twin model with direct views of reality through lenses. However this technology is a few years away.
- LIDAR is a technology that can be used to create digital twins of existing entities. Very accurate and relatively low cost.
Looking to the future, the cost of data capture is dropping dramatically (i.e. sensors). How do we synthesize, visualize and utilize this opportunity? As architects, how can we ensure future-proofing for buildings we are designing now?
When data is gathered over time within a digital twin, Artificial Intelligence (AI) and Machine Learning (ML) can be used to forecast future behavior of the system.
AI and ML need immense quantities of data, so working in siloes hinders this learning. Instead, there is immense value in sharing data across industry partners to feed AI and ML algorithms. On another hand, we recognize that while AI and ML can be incredibly beneficial, it will be critical to retain privacy. To do so, identifiers can and should be removed from data anonymize it.