AI powered Machining Feature Recognition

Updated: Aug 27, 2021

Manufacturing industries have widely adopted the reuse of machine parts as a method to reduce costs and as a sustainable manufacturing practice. Identification of reusable features from the design of the parts and finding their similar features from the database is an important part of this process.


Computer-aided process planning (CAPP) helps determine the processing steps required to manufacture a product or its parts. It serves as the connecting link between Computer Aided Design (CAD) and Computer Aided Manufacturing (CAM). Automated machining feature recognition is a critical component in the detection of manufacturing information from CAD models. When it comes to machine parts, features are semantically higher level geometric elements such as a hole, passage, slot etc. Feature recognition is the process of identifying these features from an image or a 3D model of these machine parts.


Approaching Machining Feature Recognition (MFR) as a supervised problem is not feasible for real-time applications in industries. This is because, different industries have their own data lakes to store the huge pool of 3D CAD models that include different combinations with respect to their requirements. Preparing a supervised corpus and training them to predict large number of classes is a complex and tedious process to achieve.

At AISS, Geometrical features were extracted from the CAD models using an inductive transfer learning technique using a model pre-trained with fully convolutional geometric features for the purpose of image registration. Point cloud registration is the underlying source task of this process. The number of extracted geometrical feature vectors from this process was varying with respect to the CAD models. In order to get the same number of feature spaces, later an SPP layer was introduced as the target task of machining feature recognition.




The inclusion of SPP layer improved the performance of feature retrieval. The testing accuracy obtained from the model was 86% and the top-5 accuracy was 95 % for 30 epochs. Pyramid pooling is robust to object deformations and is suitable for data of higher dimensions. Hence it can handle machining features with varying physical dimensions and it proved to be a better method than taking matrix norm of the extracted features directly.


For more technical details, you shall refer our paper.

@inproceedings{Kamal2021GeometryBM, title={Geometry Based Machining Feature Retrieval with Inductive Transfer Learning}, author={N. Kamal and HB Barathi Ganesh and V. Sajithvariyar and V. Sowmya and K. Soman}, year={2021}, arXiv:2108.11838}

We shall integrate our solution with your 3D model data bases for Machining Feature Recognition and structuring your unstructured data bases. Stay tuned, we are making similar solutions with latest AI research methods to meet your business requirements.


Unlike the traditional way of identifying business objectives and then developing complete products around those business requirements, components of AISS can be seamlessly integrated with each other in various sequences to create a complete solution which can achieve specific business requirements thereby greatly reducing associated time and costs.

If you would like our experts to provide a walkthrough of the AISS, please request a demo by filling out the form. A member of our team will reach out to you shortly to schedule a meeting.



32 views0 comments

Recent Posts

See All

Sentiment Analysis Sentiment analysis is the process of accessing the customer’s input to determine the emotions or sentiments expressed by the customer about the brand, product, and services provided

It is a technical process of splitting up audio recordings which have the multiple speakers into homogeneous segments. These homogeneous segments are related to each individual speaker. In short, it a

In the digitized world, there are multiple platforms that contain a vast amount of unstructured data (text data) that is used by companies to enhance their business and understand their customer requi