Digital Costing is a major Recurring Costs reduction enabler. A powerful analytic engine computes the target should price at part level, even in very large Bill of Material (+100 000 references)
- 10’s of Gb of data, from engineering, procurement, costing, etc.
- Multiple file format, including csv, stp, jpeg…
- The key challenge was to cross all available sources and to extract automatically the required cost-drivers (sources include 2D scanned drawings, requiring computer vision algorithms to detect specific features).
- Advanced analytics have demonstrated satisfying results on technical cost drivers extraction (Dim. Weight, material, nb of bendings, holes, welding…) from almost any source format (including scanned engineering drawings, etc)
- Parametric formula calibrated for several technologies: Machining, sheet Metal, profile, assembly, tubes & pipes, ramps… and innovative automated similarity analysis to identify spend discrepancies and target procurement efforts
- The solution was developed entirely with open source techno (Python, PostgreSQL, Tesseract, …), including a graphical interface to visualize results.
- A major ROI was achieved in a few months, with already succesfull renogociation with suppliers
- The project continues with digital costing on an wider scope, with the objective to accelerate RC reduction