Boeing Uses Stat Data to Get the Lead Out
The manufacturing sector that is leading the way in factory automation and the leading edge of the Internet of Things, the so-called industrial IoT, is increasingly relying on statistical data to develop simulation tools that could be used for new automated manufacturing processes.
Case in point is a collaborative effort between Boeing Co. (NYSE: BA) and university researchers in the U.K. focused on leveraging statistical data to improve manufacturing techniques. The University of Sheffield's School of Mathematics and Statistics announced an initiative this week with Boeing's Advanced Manufacturing Research Center to apply statistical techniques to critical machining projects.
The partners said the project focuses specifically on automating the selection of cutting parameters for machining titanium components. The goal is the creation of "time and cost efficiencies," the partners said.
The effort is critical for aircraft maker Boeing, which relies on titanium components to shave pounds off its airliners. For example, lightweight yet strong titanium components are widely used in Boeing jet landing gears.
Large manufacturers such as Chicago-based Boeing often encounter problems when machining materials like titanium because materials properties can vary from one batch to the next. (Boeing buys much of its titanium from Russia and the metal also is used in airframes and wing assemblies.)
Those variations have created a requirement for automated machining processes that account for differences in material properties. Hence, the University of Sheffield statisticians approached the Boeing center's machining group to look for ways to automate the selection of cutting parameters for machining titanium components.
"The variation in material batches not only affects dimensional accuracy and surface quality of a finished component, but also tool life during machining, all which contribute to waste and scrappage," noted Hatim Laalej, the machining group's project engineer.
The researchers collected data on machining parameters such as temperature, cutting forces and vibration to develop a statistical model that replicated the machining process for titanium. In addition, the model was used to extract data from simulations of the machining process.
The data generated from the model and cutting trials was then used to identify the optimal cutting parameters for use during actual manufacturing processes. The automated process accounted for material property variations among different batches of titanium.
"The challenge here is in how to summarize large amounts of data from multiple sensors and integrate the data with the [finite element] model predictions to get useful, usable results," added University of Sheffield researcher Keith Harris. "One aim is to identify correlations in the data to predict the average lifetime of a machine tool."
Once the optimal cutting parameters were identified, the researchers gauged wear and tear on machine tools. Sensor data was then used to develop a statistical process control approach for automating the replacement of cutting tools.
The Boeing-university manufacturing effort also underscores how industrial applications are at the forefront of IoT deployment efforts. Along with manufacturing applications, aviation manufacturers are using IoT sensors to collect data on engine performance. That information is used for applications such as maintenance scheduling.
Meanwhile, aircraft manufacturers like Boeing are essentially fielding platforms like the 787 Dreamliner that are akin to "flying datacenters."
