Participants needed for thesis research

[Minjung Ryu]

Dear Society Members,

We are continually looking for new ways of making our software products more efficient and better so that these bring more value to you.
One excellent opportunity for us doing this is our involvement in research projects with international universities or research institutions.
In some projects, we are directly involved as a company, others are directly carried out by our colleagues themselves e.g. as final theses.

One of our current projects is about to investigate how we can classifications dynamic and intelligent. The central objective is to reduce the rate of unclassified or misclassified components with the help of AI methods.
Since the current classification in SMC is based on user-defined rules, users need to manually assign the target classes or update the rules to resolve unclassified components.
To improve this, the research aims to develop an automatic classification engine for building components based on the properties.
Using the pre-assigned classification information of components as training resources, the AI system could learn and predict the matching categories for unclassified components without any manual work.

In order to get meaningful results in this project, we are currently looking for project partners, who can support us on one hand with their ideas and on the other hand with input data.
We are looking forward to use real-world classification cases to analyze and validate the performance to build a more reliable and accurate system.

If you are interested, please contact us at email, [email protected] or [email protected]

Looking forward for your contribution,

Cornelius Preidel, Minjung Ryu
on behalf of the Solibri Team

[Mats_Knutsson]

Hi Solibri,

For your development I suggest:

*looking at how the UI of Solibri could improve to leverage the strenghts (existing and new) of the program.
*adding data cleaning possibilities like SimpleBIM.

Cheers,
Mats