IPQI project's goal is to develop, implement, demonstrate, and pilot in the industry, a systematic closed-loop quality improvement methodology with in-process adaptive monitoring capable of self-recovery from 6-sigma quality defects.

As a part of this project I developed the STAS methodology, which enables the prediction of whole part deviation of a free-form surface based on partial measurement of the surface. Partial measurements reduce the time needed for the inspection of a given part and enable the effective in-line application of a 3D-Optical scanner.

The video below illustrates the methodology applied to an automotive door inner component, where, it can be seen that, the prediction error is almost zero after completion of 3 out of the total 18 measurements required to measure the entire part. A conference paper detailing the work in this topic can be found here (open access). A journal article describing the research can be accessed from here (open access) and the code to implement the STAS methdology can be accessed from here .

Non-Ideal Part Modelling

Achieving zero defects in manufacturing brings forth the need for simulations that emulate real processes and products as accurately as possible. Such simulations heavily rely on accurate representation of non-ideal parts i.e. models of actual manufactured parts with geometric and dimensional errors caused due to uncertainties in manufacturing processes.

A morphing Gaussian Random Field methodology was developed to address the limitations of state-of-art methodologies. A conference paper detailing the preliminary work in this topic can be found here (open access). A journal article is in review (arXiv version) and the code to implement the mGRF methdology can be accessed from here.