Physics-informed neuromorphic learning: Enabling scalable industrial digital twins

Open Access

Table 1

Comparison of average time consumption to generate and deploy the digital twin^a)

Processes	MDT	STGNN-DT	PINL
Model construction	1–4 weeks	0	0
Algorithm implementation	1–4 weeks	30–60 min	10 min
Model training	0	20–40 min	< 25 s
Model testing and parameter adjustment	> 1 day	1000–2000 min	< 1250 s
Total order of time magnitude	weeks	days	min

a) Here, we compare our method with the model-based DT (MDT) and the state-of-the-art machine learning-based DT (STGNN-DT).

Current usage metrics show cumulative count of Article Views (full-text article views including HTML views, PDF and ePub downloads, according to the available data) and Abstracts Views on Vision4Press platform.

Data correspond to usage on the plateform after 2015. The current usage metrics is available 48-96 hours after online publication and is updated daily on week days.

Initial download of the metrics may take a while.