Automating Airframe Measurement Improves Process Efficiency

Published: Thursday, November 17, 2016 - 12:37

Being able to be in two places at once has been a dream for many metrologists in the measurement services business. Taking a page from the field of artificial intelligence (AI), ATT Metrology developed an “Expert System” to perform a wide range of measurement tasks based on more than 20 years of hands-on experience serving customers in the metrology field. By embedding this knowledge into an automated framework, we have enabled our technicians to be in not just two, but many places at once.

The project began when a large aerospace customer needed a method to measure airframe modifications before and after the modifications were applied. The measurement process needed to allow field technicians to perform measurements on an as-needed basis. The process also had to ensure accurate data collection and produce useful information to the engineering and design staff in charge of the modifications. The ATT team proposed and delivered a dedicated software program to address the customer’s airframe measurement needs, which provides real-time data feedback on location and ensures measurement quality.

Introduction

As measurement service providers, we have seen rapid technological progress in the field of metrology over the last 20 years. The level of measurement and analysis we are capable of today was once only possible in the laboratory and under the supervision of scientists and engineers. Today, the tools have been improved and streamlined with the use of third-party software to do much of the analytical heavy lifting. This progression has allowed a generation of metrologists to apply their skills and discover best practices in a wide range of applications. These advancements also led to the formation of organizations such as the Coordinate Metrology Society to capture and disseminate standards and practices for a whole new generation of metrologists.

The natural progression of metrology technology is making it possible now for non-metrologists and metrologists alike to perform complex measurement tasks and producing useful results using software that encapsulates industry best practices, process knowledge, and metrology.

Project overview

The hardware used on this project consists of a Leica Absolute AT-960 Tracker with a 6DoF Leica T-Probe.

Figure 1: AT960 and T-Probe

Using a dedicated user interface (UI) with audible and visual cues, the Airframe Monitor application guides operators so they can complete measurement jobs quickly and accurately with a minimal required background in metrology or laser tracker experience. The application interfaces with SpatialAnalyzer software to monitor the location of a Leica T-probe relative to the airplane being measured and provides direction to the operators on where to locate the desired points of interest on the airframe as the operator approaches them. The program collects and verifies measurements in real time and ensures only good data is captured.

Figure 2: Airframe monitoring application. Click here for larger image.

Data validation is performed against the nominal data as well as the measured point data and measurement statistics, such as min/max RMS. The program provides modules for performing on-site operation checks for two-face and ADM validation. In addition, it ensures that scale, drift and the environmental properties are properly performed and logged. It permits the movement of the laser tracker to different locations and automatically guides tie-in back to the previous station locations. Once the required measurements are completed, the program ensures any required analysis, such as best-fit analysis, is automatically initiated so that technicians can generate and examine reported results.

The points to measure on the airframe are a combination of fastener holes, surface points, and scanned planes. The large number of points, features, and their locations makes it prohibitively difficult for the technician to visually identify these features by location or by memory. 

Figure 3: Points grouped into selectable locations. Click here for larger image.

For this reason, it is crucial to have the software provide a proximity guide to the operator during measurement. We use a visual display on a large LCD monitor located near the operator workstation to show key location information and a graphical proximity indicator. In addition, an audible proximity tone guides the operator to hole and surface locations without need to constantly monitor the display.

Figure 4: Measurement guidance display

Another important feature to enhance the speed of measurement collection is the automatic identification and naming of collected points so they may be measured in any order (as seen in figure 5). This allows the operator to measure points in the most convenient order. Points that are skipped, or that are out of the line of sight of the current tracker location, can be picked up later from a new tracker location. Tracker locations can be added as needed.

Point data are placed in a date- and time-stamped collection. The baseline data can then be compared to future measurement session data. The comparison report is saved in the same collection. Reports take a few minutes to generate and can be reviewed immediately after the measurement session. This allows operators to ensure that the data have been collected successfully and that reports look good before breaking down and storing the equipment.

Figure 5: Point naming, matching, and fitting

The automation is designed to interpret a job template that includes nominal point data for certain aircraft sections and is instrumented with other data about the job (i.e., pictures, point groups, and views named according to section, etc.). This type of configuration “by convention” provides a logical structure that a program can easily work with. This allows us to adapt the program by changing the job template. In theory, the process can be applied to any set of points we wish to monitor and compare. The job template is organized using conventions recognized by the automation engine. In this way, future templates can be developed without the need to change the core software.

Outcome

Based on past experience with similar jobs, we estimate that the same job done manually by a trained technician familiar with the process would take 8 to 10 hours of work. In actual measurement projects using the airframe monitoring program, the work is taking only 2 to 4 hours depending on the quantity of data collected. This amounts to a conservative estimate of time savings of 50 percent or more over manual collection, with a much better consistency and accuracy of data collected.

Ease of use was a key design goal as it reduced the amount of training required to run a measurement job. Documentation and interactive graphics in the application prompt operators for actions and inputs as they are needed. With a short set of orientation videos, operators are instructed how to setup and check the equipment and then start the software. We tested usability by having new operators, with only a few minutes of instruction, perform a measurement job while collecting their feedback on the ease of use. 

Since delivery of the system, ATT has improved efficiency of the process in several ways:
• Reduction in time by at least 50 percent in collecting data over manual methods.
• No longer needing a human assistant to guide the technician and address beam interruptions and software messages and dialogs during a measurement job.
• Tasks related to data management (i.e., point naming, incorrect point deletion, etc.) are automatically handled and addressed so that technicians can work with clean data instead of wasting time fixing data. 
• The application instantly generates standardized reports for the engineering staff to evaluate.

About The Author

Michael O’Shea

Michael O’Shea joined ATT Metrology as chief technology officer in April 2015. He brings more than 25 years of experience building and leading the development of software solutions with applications in a wide variety of fields such as HR, aerospace, and financial services. Most recently he was manager of software development for Russell Investments. Prior to joining Russell in 2007, he was western regional director of software development at Logicalis serving a long list of clients in telecommunications, biotech, and a number of Seattle internet startups. He was a software engineer at Boeing for 10 years from 1989 to 1999.