SBIR Phase I: Artificial Intelligence (AI)-Aided Part Identification for Coordinate Measuring Machines

小型企业创新研究(SBIR)第一阶段项目的更广泛影响是开发用于测量零部件和组件的新一代智能机器。该团队已经证明，这项技术可以将现有的坐标测量机转换为自动驾驶机器。自动测量零件的能力是工艺链中的一个重要反馈环节，它将使未来的制造完全自动化。具体地说，这种自动化将降低使用坐标测量机(CMM)所需的专业技能。这项创新将使工人能够操作三坐标测量机，并获得精确的零件测量。这种设备特别有用，因为熟练的制造/计量劳动力即将退休，因为它使新员工能够提供准确的信息，而只需很少或根本不需要培训。这一创新还为制造公司提供了购买新机器或升级现有坐标测量机的选择。虽然这项提议的重点是零件识别，但这项技术在计算机数控加工、机器人和自动化装配线上已经有了现成的应用。这一能力将使美国制造业变得更加强大和技术更加先进。这项提议的目标是开发一种新的技术来识别机械加工的部件和组件。这项技术将在三坐标测量机(CMM)上实施，三坐标测量机在制造业中广泛用于测量零件的形状和尺寸。这项拟议的技术将实现零件的自主测量，从而实现更高水平的自动化。在这项识别技术中，该团队将使用来自相机的实时图像、多个实体模型/计算机辅助设计(CAD)生成的图像以及高级图像处理。将人工智能(AI)/机器学习(ML)应用到零件图像处理中，可以保证零件的正确识别。摄像机所看到的零件的正确识别是实现零件的自动测量的剩余未解决的挑战。大多数加工零件是无纹理的，大部分信息都包含在边缘中。目前的图像处理技术对纹理丰富的零件效果很好，但对于无纹理的已加工零件则不可靠。使用边缘和形状信息的AI/ML增强图像处理是一种很有前途的方法，解决这个问题将导致新一代可以自动测量零件的坐标测量机的诞生。该奖项反映了NSF的法定使命，并通过使用基金会的智力优势和更广泛的影响审查标准进行评估，被认为值得支持。

The broader impact of this Small Business Innovation Research (SBIR) Phase I project is the development of a new generation of smart machines used in the measurement of parts and assemblies. The team has demonstrated that this technology can convert existing coordinate measuring machines to self-driving autonomous machines. The ability to automatically measure parts is an important feedback link in the process chain that will enable fully automated manufacturing of the future. Specifically, this automation will reduce the specialized skill required to use a Coordinate Measuring Machine (CMM). The innovation will enable workers to operate a CMM and get a precise part measurement. This device is especially helpful as the skilled manufacturing/metrology workforce is retiring as it gives new employees the ability to provide accurate information with little/no training. This innovation also gives the manufacturing companies an option to buy a new machine or upgrade their existing coordinate measuring machine. While the focus of this proposal is part identification, this technology has ready applications in Computer Numerical Control (CNC) machining, robotics, and automated assembly lines. This capability will make the US manufacturing sector stronger and more technologically advanced.The objective of this proposal is to develop a new technology to identify machined parts and assemblies. This technology will be implemented on coordinate measuring machines (CMM), which are used widely in the manufacturing sector to measure the shape and size of parts. The proposed technology will enable autonomous measurements of parts allowing a higher level of automation. In this identification technology, the team will use live images from a camera, multiple solid model/Computer Aided Design (CAD)-generated images, and advanced image processing. Applying Artificial Intelligence (AI)/Machine Learning (ML) to the image processing of part images will ensure correct part identification. Correct identification of parts as seen by the camera is the remaining unsolved challenge to achieving self-driven automatic measurements of parts. Most machined parts are textureless and most of the information is contained in the edges. Current image processing techniques work well with texture-rich parts but are unreliable with textureless machined parts. AI/ML enhanced image processing using edge and shape information is a promising approach, solving this problem will lead to the birth of a new generation of CMMs that can measure parts automatically.This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.