**Description**
The control of automated riveting machines in the aerospace industry have undergone several major evolutions since they were first introduced. Nearly all machines in use today are driven by part programs produced by processing 3D CAD data in specialized offline program generation software. However, before this became standard practice, machines were designed to have the operator manually drive the machine to align a set of camera crosshairs on the fastening point. This manual process also involves a pre-machine production step to lay out mylar on the panel and mark each fastening point with a colored dot that corresponds to the fastener type and size. This whole flow is extremely labor intensive and introduces many potential sources of human error into the production process.
As part of a contract to replace one of these aging manual machines, Electroimpact was also asked to provide a “teaching” function that would allow the creation of a part program by using manual inputs to the machine. Manual teaching for industrial robots is a mature technology, but we are not aware of another case where it has been applied to a large-scale riveting machine. This presented us with some unique requirements and challenges to overcome. Electroimpact’s solution has resulted in a fully integrated system that collects the teaching data fully automatically with very limited additional interaction required from the operator. Our customer can then import the collected data into custom-built program creation software which provides a 3D visual interface of the manual commands and outputs an NC part program allowing fully automatic production of that same part. Before the program is output, we have also provided tools allowing further optimization of the program beyond what is possible with manual input.
Additionally, as the machine must still be run manually to generate the teaching data, Electroimpact implemented additional safeguards against errant operator commands. A vision based tool was created to check that the machine is positioned correctly over the painted dot and that the correct cycle conditions are selected for the dot’s color.
Our solution to machine teaching de-skills the teaching process, reduces human error, and will ultimately allow our customer to produce parts at higher rate and with far fewer defects than was possible with their existing equipment.

**Author(s) - ranked in descending order of contribution / as it should appear when published**
Isaac Bahr - Director & Lead Engineer, Electroimpact Japan K.K.
Conor Durdy - Senior Machine Vision Engineer, Electroimpact Inc.



De-skilled teaching for large scale riveting machines

SAE International’s AeroTech is the premier global aerospace technology conference and exhibition focused on and celebrating innovation in virtually every facet of modern aerospace and defense mobility, including platform, system, subsystem, component, and material research and development, engineering, manufacturing, assembly, test and validation, certification, integration, deployment, sustainment, safety, security, and sustainability. Innovators and thought-leaders from the global aerospace community convene annually at AeroTech to share and learn about the latest advancements and achievements, advance technical knowledge and innovation, discuss challenges and practical solutions, discover new technologies and techniques, uncover trends and opportunities, network, and collaborate.