Automation of the Quality Control Process with the Use of Robotics and a Coordinate Measuring Machine

Alexander Hoang, Purdue University

Abstract

Micro, small, and medium enterprises (MSMEs) are the corner stone of the modern economy accounting for 99% of all businesses and over 50% of all employment in the world (Sariyer, 2021). Governments in countries, like India, that understand the importance of MSMEs give incentives for additional MSMEs to be established and the existing ones to flourish (ASSOCHAM Bulletin, 2016). India have been developing methods to boost the quality of workers that work in MSMEs so better products would be manufactured for domestic and international use. Despite the importance of MSMEs, they have struggled to maintain a competitive advantage in quality control compared to their larger competitors (Sariyer, 2021).Quality control continued to being an important part of the production process to ensure that the manufactured products meet all the criteria defined by the customer. The quality control process traditionally involved an inspector acquiring random samples from the production line as the product was being made or acquiring the completed product at the end of the line (ASQ, 2022). The importance of performing quality control and quality assurance checks was to ensure quality products are manufactured. Lean manufacturing was a fundamental ideology for manufacturers where they focus on maximizing productivity and minimizing waste simultaneously. When it came to production, manufacturers ideally wanted to have their process fall within the methodology of six sigma. Six sigma was the idea that for every million parts, only 3.4 parts were scrap (Sower, 2011, pg. 59). Both approaches combined created the method known as lean six sigma, a method where companies tried to maximize production, reduced the amount of scrap, and made an average of 3.4 scrap parts per million parts. To work towards this goal/methodology, companies needed to implement quality management systems (QMS) to ensure that they were following the steps needed to improve the quality of their product.Despite the move towards the automation of the manufacturing process, there was lack of development in the automation of the quality control process for in-depth inspections like those done by a human inspector. Quality control was an area of production that needs automation due to the subjectivity of human inspection (Taranito, 2022). Some modern end-effectors had the ability to inspect features on parts, but they forgo certain features in quality control when compared to a coordinate measuring machine (CMM). Laser vision end-effectors were the newest endeffectors that was potentially used for quality control. Laser vision end-effectors can get features like the shape and surface roughness of a hole easily by scanning the feature. However, they typically required the part to be stationary as well as requiring the end-effector to be clean to get a clear scan of the feature (Tomov, 2017, pg. 484).A camera was another common way to check features on a part, but the issue was that sorting algorithms need to be devised so the robot knows the difference between defective and non-defective parts. The feature needed to be noticeable enough for the camera to detect if something was wrong on the feature (Morey, 2017, p. 69). The problem was a lack of a system that automatically inspected a part near the same precision and accuracy of a CMM without having to stop the production process. Robotic arms have been used to load CMMs, but only if the part was placed into a storage unit by a human operator. MSMEs struggle to implement systems due to expensive costs, sophistication, complexity, and space requirements. The Hexagon TEMPO attachment for their CMM, shown in Figure 1.1, was an example of a complex system that does not allow seamless integration.

Degree

M.S.

Advisors

Gan, Purdue University.

Subject Area

Industrial engineering|Robotics

COinS