A Flexible Manufacturing System Architecture for On-Demand, Rate-Independent Production

Abstract

For low to medium production volumes or products with short life cycle, dedicated manufacturing lines often end up being a commercial burden. The high initial non-recurring expenses of the manufacturing infrastructure are eventually passed onto the cost of the product, limiting its commercialization potential. In such cases, flexible frameworks for manufacturing are ideal, at least in theory, with reusability of the same hardware for different product developments. From an implementation standpoint, however, flexible manufacturing platforms are difficult to deploy as they suffer from inherent uncertainties that severely constrains the overall precision in task execution. Low cost flexible robotic manipulation systems, configured using individual modules from an assortment of positioners, often lack the precision required for assemblies with stringent tolerance budgets and therefore have to rely on expensive feedback systems for reliability at the trade-off of process throughput. In order to overcome this fundamental challenge with flexible or modular assembly and packaging systems, this paper presents a novel precision evaluation and control technique. The proposed solution estimates and tracks the end-effector position errors in a robotic manipulation system resulting from the kinematic configuration as well as the dynamic parameters for each specific task; thereby, allowing the automation application to compensate for these errors in run-time. The computation does not require active feedback, thus ensures high throughput while maintaining high reliability. The experimental results provided in this paper demonstrate the efficacy of the proposed approach in the case of a high precision micromanipulation task using a custom robotic manipulation system. The analysis also demonstrates that different system configurations offer different levels of precision metrics.