Infographics of the i-feature intelligent system for processing standard elements in detail

    Yesterday I made a small infographic of my development (read: habrahabr.ru/blogs/artificial_intelligence/135795 )

    I'm sorry for unprofessionalism - I'm just learning to design in this direction. I have long wanted to start somewhere.

    I present the prospect of developing my future system in the context of using Cloud technologies and data storage and processing systems.



    I explain briefly under the cut.



    It is divided into three parts:
    1: collecting data from users and transferring them to servers for processing. There, training and test samples are formed. The sample for retraining is also updated there (or the parallelism of two neural networks is used for more optimal and operational retraining)
    - Separately, it is worth noting the development of prospective means of analysis of the solid-state model, i.e. checking for thin-walledness, for surface curvature, etc. For some reason I’m sure that this is real, because each surface in the model has its own mathematical description in space. This data also goes to the clouds to train the corresponding neural networks.

    2: Actually developed technique. In this case, the "i-feature" works on the principle of client-server. Those. all mathematics is taken out into the clouds, the subroutine receives answers and sends requests. The linear algorithm works similarly.

    3: Stage of final testing by the user of the obtained results of the algorithm. Everything is reduced to a MINIMUM of parameters, which greatly simplifies the work with the system (especially for beginners). The tool is either created parametrically (by the way, it came up with how to integrate the neural network there), or it takes from the already created library of the tool. It is also worth noting the very real possibility of creating a single tool library for various manufacturing companies! A kind of globalization. There are no special difficulties there - mostly mechanical work. The tool parameters in the library can be used to train individual networks, which will significantly help with the parametric creation of the tool!

    Further, by clicking on a couple of “OK” buttons, everything goes into the standard CAD mode (generating trajectories according to the approved parameters and postprocessing).

    As promised, I’m conducting work, constantly improving the idea and algorithms.

    Thanks for the feedback in the previous post!

    Also popular now: