4D Point Cloud Data Processing, Feature Extraction and Predictive Modeling

Our primary objective is to develop robust and accurate predictive model that utilizes the 4D point cloud data to predict product properties for future products/objects. The model will enable us to enhance our product development, quality control, and overall decision-making process.


P&G has accumulated a vast amount of micro-CT scan data. The data has been processed into 4D point clouds (x, y, z, d) data consisting of 50K to 60K points per product/object.

Description of the need:

We are seeking consultations with experts who possess in-depth knowledge and experience in 3D point cloud data processing, calculation of geometric features and AI/computer vision prediction.

What we are looking for:

We are seeking consultations with experts who possess in-depth knowledge and experience in the following areas.

  1. 3D point cloud data preprocessing

    • Expertise in preprocessing methods for 3D point cloud data, such as orientation alignment, denoising, and normalization.

  2. Calculation of Geometric Features

    • Deep understanding of methods and techniques for calculating both local and global geometric features from 3D point clouds.

    • The ability to provide python codes that efficiently extract geometric features from point cloud data.

  3. AI/Computer Vision Prediction:

    • Expertise in innovative computer vision techniques and AI algorithms to predict product properties using 4D point cloud data obtained from micro-CT scans.

    • We have some preliminary approaches and are looking for new and better approaches.

We are not interested in:

  • We are interested in techniques implemented using Python (no other software).

  • We do not require individuals to perform data preprocessing, modeling, or programming task.

Please note that only non-confidential information describing the business & services model, current use and IP can be accepted for review.