Digitally Assisted Assembly at Factory 2050
In a previous article, we introduced the University of Sheffield’s Advanced Manufacturing Research Centre (AMRC), a member of the AREA that develops innovative techniques and processes for high-precision manufacturing. A subsidiary, the AMRC with Boeing, collaborates with a variety of research partners in areas such as informatics, automation, robotics and Augmented and Virtual Reality. Besides aerospace, the results of this research into manufacturing are used in the automotive, construction and other high-value industries.
Earlier this year, the AMRC opened the doors of its newest manufacturing facility, Factory 2050, a glass-walled reconfigurable factory in Sheffield Business Park. The facility investigates and showcases new technologies and processes relating to Industry 4.0, including projects to explore digitally assisted assembly technologies to fill a looming skills gap in the aerospace industry.
Augmented Reality in Digitally Assisted Assembly
The Digitally Assisted Assembly (DAA) project focuses on techniques for delivering work instructions to factory operators, including the use of optical projection AR and wearables. According to the AMRC’s digital manufacturing specialist, Chris Freeman, the project allows partner companies to experience visual work instructions through a number of delivery mediums. Research includes:
- Optimizing AR tracking methods for effectively getting a part’s position to generate a frame of reference.
- Designing user experiences for work instructions that are projected or overlaid onto a part within the user’s field of view. These include instructions that guide users for tasks such as gluing sequences, fastener insertion, inspection, wiring looms, complex routines and more. The aim of this research is to reduce cognitive load and optimize the user experience for delivery across a variety delivery modes (e.g., projection AR) and devices from tablets to smart glasses.
- Using location-based services to add contextualized task and environmental information in relation to the user’s position or progress within a task.
With the technology still in its infancy, one of the aims of DAA is to simply demonstrate what can be achieved with the technology. Although smart glasses and wearables aren’t proven or certified for use in manufacturing, they are nevertheless being baselined for further research and possible future production usage. The AMRC are currently following a strategy of first identifying the “low-hanging fruit” from the current state of hardware and software, which means that research associates want to find some of the most obvious and perhaps least expensive options up front.
Skype for HoloLens
Although the AMRC is studying a variety of smart glasses brands such as ODG and Vuzix, remote collaboration use cases with Skype for HoloLens is an interesting application for meeting the needs of certification processes. This use case includes methods for lineside support and remote verification to complement or replace expensive quality management activities requiring the presence of a supervisor. It may even include assistance by remote colleagues when assembly or repair problems are encountered.
Freeman notes that though such use cases aren’t spectacularly advanced in terms of tracking in comparison with other scenarios such as overlaying geometric 3D models on objects being assembled, they are nevertheless disruptive of current manufacturing practices.
Projecting Work Instructions on Large-Volume Objects
Projected Augmented Reality, sometimes referred to as “spatial Augmented Reality,” features one or more optical projectors projecting a beam of light onto a specially designed work surface or even on the parts being assembled. Thus work instructions are displayed directly on surfaces to guide operators. The DAA is currently researching methods for effectively using projection AR in conjunction with both fixtures and robotic arms in work cells.
For example, an operator assembles aircraft parts with the assistance of robots to present a part at a better angle than if it were lying on a work surface. A robotic arm can swivel or position the part as needed by the operator, and projected AR is able to guide operators through a series of specific manufacturing procedures.
As has been discussed in other industry contexts, return on investment on any new technology can be challenging to define (whether it’s for AR or any other). Typical ROI calculations seek to determine the amount of savings a project can bring and when that investment will pay off. In the case of AR, relevant questions include how to quantify the value of conceptualized data and geometries for its usage in performance metrics.
Further research into AR will eventually uncover such answers, but in the near term, human factors and ergonomic studies can also quantify the technology’s effectiveness. For example, the AMRC is currently conducting AR-related training scenarios to determine a variety of metrics such as memory retention and AR’s overall effectiveness, as well as usability and operator response.
Although research being conducted at Factory 2050 aims to advance the state of the art in aerospace manufacturing, many of the techniques and procedures derived by DAA and other projects will eventually be used in other industries, such as energy and construction. For example, assembly techniques for large-volume aerospace parts can also be applied to assembling prefabricated homes at a factory as part of modular building manufacture. Having recently opened its doors, it’s apparent that the new facilities of Factory 2050 will have an impact on both present and future manufacturing in multiple domains for many years to come.