Remanufacturing - Utilising machine learning to automate/assist in part restoration.
Background: Currently, products reaching their “End-of-Life” (EoL) that still hold value in refurbishment end up being rebuilt with new parts rather than the damaged parts being repaired/remade, the damaged parts being landfilled or recycled for their material (losing all value from its original manufacture). This is, for the most part, solely because it is an easier, cheaper, and less intensive solution.
With global resource shortages and increased costs to find, procure, process and ship, making remanufacturing a viable process can make an incredible impact globally, in the drive to become more sustainable and save diminishing resources.
The research will primarily centre on utilizing industry data to develop the programming algorithms necessary to process imagery to identify and categorize damage and failure modes, and then construct a remedial plan. From that plan, the software would then need to generate the instructional toolpaths (gcode) to operate the multiple tools required to fix the damage.
The final goal is to have the software tailored to the company using it, allowing them to quickly and accurately locate, diagnose and repair damage to their machinery/products, on a time scale and budget that makes the process financially beneficial.
Current efforts are developing my Python and coding skills through LinkedIn Learning - as this will be the language used throughout my research.
Realistically, this development will continue well into the official start of my PhD in October. After which the main aim will be directed at the development and testing of the software.
