-
A period of reflection…….
A very sunny and warm hello to fellow enthusiasts !!!
I have been reflecting on the quite unbelievable Digital Twin journey over the last 2 years and having to pinch myself at the progress that has been made.
Within Costain, Digital Twin is recognised in all corners as something our industry needs to do, in the big wide world the same can also be said with what seems to be an ever increasing global drive!
Wanting to challenge not just myself but the community at large, here are a two ‘big ticket’ items of reflection where it feels as though we haven’t yet fully succeeded:
1 – Collaboration – now don’t get me wrong, the collaboration from DT has been truly exceptional and is, I believe, changing industry. However are we still creating digital dots that are yet to be properly connected?
For NDT to really be successful we need to reach a point where open conversations can take place between Gov, academia and industry in a way that has not traditionally happened. Discussions that are ‘warts and all’ that would not normally happen between say an owner / operator and supplier, or between Government and industry.
 We risk restricting the value of DT which will verge on the magical if truly transparent.
2. – Engineering mindset – No surprise here that as an engineering led industry the focus with DT’s appears to be largely engineering focused.
In a recent piece of work looking at data trusts, the legal complexity of scaled data sharing has been eye opening. What if a decision is made based on data from say 10 organisations, and that decision leads to an issue because of some low quality or incorrect data, who is then liable? Would it be possible to identify the bad quality data?
Exposure to leading research has identified the complexity in privacy, ethics, trust, reliability, accountability and security in relation to collaborative data sharing.
It was also interesting to hear about work by the Financial Conduct Authority looking at what a Data Conduct Authority might look like, where data might be monetised.
As we look past the engineering foundations, there is a lot to do.
I hope people do not view any of the above negatively, this all screams opportunity and is only natural as we lead the world in the development of scaled, federated Digital Twins.Â
What other big challenges do people think need some focus?
Regards
Kevin
-
2 Replies
-
Hey @DRossiter87 – thanks for sharing. I thought this is a good report. Since I saw your post here, I’ve been on a research exercise to find more references to “cost of bad data”. Have you come across other studies that assess or discuss that cost? I think it is helpful knowledge to help assess the value you can get from data; both value of static data and and value of data flow between entities (like within an ecosystem of connected digital twins).
The other reference for bad data cost I came across recently is:Â https://hbr.org/2016/09/bad-data-costs-the-u-s-3-trillion-per-year
-
Hi @DRossiter87, there is also a gov article linking back to the framework (ps your link seems to have gone stale).
The magnitude of datasets seems to make writing the rules as hard as creating the data and the evolving nature of the Employers Information Requirements creates a dynamic validation scenario.
The requirement to validate data is not only at the schema level, but the downstream consumers. Â
I highlight some of the challenges with reviewing datahttps://medium.com/@allowing_pullman_wasp_147/validation-of-property-instances-with-itwin-c04982134f52
and assets such as panels for example prevent the rapid deployment of engineering content into operations.
I have various threads running on how to validate, but they all seem to be asset owner / project specific.
The only government standard I seem to have come across is the Singapore Corenet system which is very specific.
Interested in what others are doing to standardise the data validation requirements.
Log in to reply.