A University of Texas at Arlington computer scientist is developing a portable computer vision system that integrates augmented reality into aircraft maintenance training for the U.S. Air Force.
William Beksi, assistant professor of computer science and engineering, has received a $450,000 Small Business Technology Transfer grant from the Air Force Research Laboratory to fund his project. He's trying to assist Air Force efforts to keep its aircraft in peak operating condition as the military branch attempts to replace the collective knowledge and skills of retiring service members.
"There are lots of manual tasks that aircraft maintainers must do each day, and we're not at the point where they can be fully automated," Beksi said. "But if we give maintainers technology, they can be much more efficient in completing these tasks. We hope our solution will significantly expedite the Air Force's maintenance training program and make it more effective."
Beksi plans to use field-programmable gate arrays (FPGAs) to make a small device that can be attached to a belt and then connected to a head-mounted display that will help the maintainer detect and identify parts in 3D. FPGAs are integrated circuits commonly used in applications where speed and flexibility are needed. They are lighter and more portable than a traditional central processing unit and they require low power so batteries will last for hours. Additionally, they can run deep neural networks quickly without lag.
A typical use for this technology could be the assembly and disassembly of an aircraft wing, where the technology would help the maintainers determine where to start and show which objects they are looking for.
"The Air Force wants to see this work because now they're tied to a specific location, such as a hangar, and their maintainers can't move around the base freely," Beksi said. "The major advantage to this solution is that its portability removes constraints so people can move from place to place and complete their tasks efficiently."
Beksi is working with the company krtkl to develop the hardware while his lab develops the artificial intelligence that will run on it. They will test the prototype at Robins Air Force Base in Warner-Robins, Georgia, with an eye on successfully commercializing the technology for use by commercial airlines or other large organizations.
"The Air Force's desire to increase efficiency in training its personnel to better work with maintenance and manufacturing tasks provides the impetus to create something that has far-reaching effects," said Hong Jiang, Nedderman Professor and chair of UTA's Computer Science and Engineering Department. "This program is an excellent opportunity for UTA to collaborate with small businesses to develop solutions that will help industry in many fields."
The Small Business Technology Transfer program is highly competitive and encourages domestic small businesses to engage in federal research/research and development with the potential for commercialization. The program relies on partnerships between small businesses and nonprofit research institutions, with the small business required to formally collaborate with the latter in Phases I and II. The program's most important role is to bridge the gap between performance of basic science and commercialization of resulting innovations.
