Quality Performance Improvement through Robotic Process Automation in Rail Manufacturing

Sustainable quality performance is key to railway manufacturing success. The paper investigates the relationship between quality Defect Per Unit (DPU) generated from a non-automated production process, vs. the same but partially automated (conceptual) production system built around Robotic Process Automation (RPA) principles, focusing on local South African railway manufacturer Company AB. Desktop literature speaks to a positive correlation between process automation and project quality performance. DPU data from South African Company AB, captured across a sample size of 20 locomotives, over a 2 to 3 year period is dissected quantitatively to establish (a) quality performance (DPU based per unit) without RPA and (b) the theoretical quality performance of the same manufacturing system after the incorporation of RPA. Process automation design and limitation of the value chain is derived from a conservative automation scale approximation assumption (benchmarking of automated process industries). The paper finds that a conceptual automation model of Company AB's value chain yields quality performance improvements of up to 65%, along with a reduction of the Cost of Non-Quality (CONQ) from 4.1 to 1.21 million ZAR. In addition, downtime is reduced from 58 to 16 days before and after RPA respectively. The result reinforces historic sentiments where RPA is generally accepted as a vehicle to accelerate process efficiency and improve overall product quality.