Georgetown College. Georgetown University.

Imagine your job is to inspect the quality of an automobile assembly process. You must check many welded joints and insure the proper fit of parts. How do you make sure you don’t overlook anything?

When mistakes could cost the company hundreds of thousands of dollars in delays and recalls, how can manufacturers insure accuracy with such a repetitive tasks?

In the automobile industry, assembly workers must develop highly refined skills to perform detailed assembly and inspection tasks. They may need to inspect over 40 weld joints in a short amount of time. Traditionally, training a new worker to obtain these skills can take a long time, and repetition can lead to inaccuracy.

Dr. Brian Blake of the Computer Science Department has developed a way to mitigate these concerns: computer-assisted goggles, or “augmented reality devices.”

The goggles work by enhancing what the worker sees, overlaying the actual visual field with a virtual map of key areas to investigate. This breakthrough technology brings the innovation of virtual reality into the real world, in real time, allowing simultaneous practical application.

With the augmented reality goggles, workers will have the ability to see computer-based images as well as the actual pieces they are working on. Blake and his collaborators anticipate that their creation will dramatically enhance on-the-field efficiency and accuracy.

If you’re responsible for inspecting the quality of assembly, the goggles will ensure an accurate and complete inspection of parts by visually tracking what joints have been checked and what joints are left to inspect.

In the factory setting, the goggles will be equipped to interact with the day-to-day needs of the worker, responding to voice-activated requests for enhanced reality assistance. They will also serve as an important assistant for workers taking over an unfamiliar job or completing a task for a sick coworker. Comprehensive real-time, audio, and visual instructions will enable efficient and thorough completion of new or unfamiliar tasks.

Developed to be an “intelligent,” interactive device, Blake and his collaborators intend the goggles to have features that enable them to tailor instructions to their user’s learning type, ideal for the training of new workers.The goggles will be programmed so that the mode and level of instructions can be adjusted to a particuar worker’s learning type and capability, decreasing cues and commands as the user gains proficiency. In addition, the enhanced reality feature can simulate the noisy, busy atmosphere of the factory previously unavailable in training.

Support for the device comes from the National Science Foundation (NSF) and the BMW car company. BMW and other car manufacturers hope that these devices will speed up processes and increase accuracy for their factory workers.

The eventual target user for the device, car manufacturers, is identified in acronym for this project: CAARS (Context-Aware Agent-Supported Augmented Reality System). Dr. Blake and his undergraduate assistant Jerome Butcher-Green collaborate with a small company, Juxtopia, LLC, with funding from the National Science Foundation Small Business Technology Transfer (STTR) program.

As additional support for the project, Butcher-Green, together with another student, Robert Browning, won a $7,000 stipend from the Collaborative Research Environment for Undergraduates in Computer Science and Engineering (CREU) program, which is sponsored by the Computing Research Association.

Print Article