Fill out the form to keep watching
Having trouble viewing the video?
Some regions or networks may restrict Vimeo playback.
The shift: Every MIDAS GEN NX operation an engineer clicks through can be driven from outside the interface
In the session
- Connect, sync and run analysis through the API
- Support reactions extracted in one pass
- ACI 318M-14 footings sized and written back as meshed plates
- AI-generated code reviewed before use
Stronger outcomes in MIDAS GEN NX workflows begin with improving how support reactions are extracted and returned to the model as designed elements.
Your submission has been
successfully received
If you meet the certification criteria,
Your Certificate of Completion will be sent to your email.
If you need urgent support, contact us via
here.
We truly appreciate your participation!
Open Access Has Ended
The open access period has ended.
If you have a MIDAS account, sign in
and complete the access form to watch this session.
Full access is one step away 👀
Complete a short form to continue
Description
| In this session | |
| 00:00 ~ 02:09 | Introduction |
| 02:09 ~ 04:11 | Running an AI agent inside the IDE |
| 04:11 ~ 12:26 | Building the prompt |
| 12:26 ~ 18:13 | Plan mode and reviewing the plan |
| 18:13 ~ 26:50 | Reviewing the generated code |
| 26:50 ~ 32:11 | Syncing, analyzing and extracting reactions |
| 32:11 ~ | Drawing the footings back into the model |
| Part 2 of 2Part 1 covers generating a model in Python through the same API. | Watch Part 1 → |
Everything MIDAS GEN NX does in the interface is also reachable from outside it. This part shows how far that goes on an existing mid-rise building model with 47 supports: a custom tool connects to the running session with the MAPI key, syncs the model, runs the analysis, and reads service and strength support reactions back out — the same operations an engineer would click through, now driven from one screen built around this particular task.
What sits on top of the API is a small program written with an AI coding agent inside the IDE, prompted with a role, a numbered list of the required functions, and a link to the MIDAS API Python documentation so the agent uses real function names. From the reactions MIDAS GEN NX returns, it sizes isolated footings to ACI 318M-14 for one-way and two-way shear at an allowable bearing pressure of 150 kPa, then writes them back into the model as plates meshed at 0.5 m with the support node kept on the mesh. Results leave the model and the design comes back into it, without a manual hand-off.
Reviewing the generated code is treated as part of the workflow rather than an afterthought. The walkthrough shows the foundation engine written for uniaxial bending where biaxial was required, plus two runtime failures — reaction extraction on load combinations, and the drawing step — each fixed by feeding the error back to the agent.
Key Points
Reaching MIDAS GEN NX operations from outside the interface through the API
Running analysis and syncing the model from a custom tool
Extracting service and strength support reactions in one pass
Sizing isolated footings to ACI 318M-14 for one-way and two-way shear
Writing the designed footings back into the model as meshed plates
Reviewing AI-generated engineering logic before it is trusted
Speaker