Mr. Schiefelbein, from today's perspective, what are the limits to the use of smart glasses?

Schiefelbein: One major hurdle is still the occupational safety certification, which ensures that the data glasses used can be used on the basis of an individual risk assessment. Also, when purchasing AR data glasses, key success factors for implementing 'real' AR are still not taken into account: the necessary infrastructure - i.e. backend integration and indoor navigation -, digitally available content and the 'right' devices for the selected use cases. Ultimately, progress on the store floor can only be achieved with the right use cases, taking the aforementioned criteria into account.

Where specifically do you see a need for further research - both technically and organizationally?

Schiefelbein: For example, in the area of object recognition and tracking or indoor navigation. There are various methods for object recognition and tracking and a very careful selection or approach is required to combine the methods. In particular, the required comparison/reference approaches - such as a photo/CAD model in combination with SLAM/edge detection - are crucial to success. If necessary, control from an upstream 'system over the air' - i.e. wireless instead of wired as in the past - is important, as geometrically identical objects may have to be handled in a work instruction, for example.

As far as in-house localization is concerned, it is currently still unclear which technical approach makes sense in combination with AR solutions. For example, beacons versus object recognition and this in turn in comparison to point clouds. The decisive factor is a solution approach that is as cross-platform as possible with an accuracy in the centimeter range. A marker-based approach may also be possible. The integration into a SmartGlass and provision of location data is still the subject of research.

In both areas, there is also a need for organizational and legal clarification - for example, regarding the recording of people by the camera sensors or the mapping of business premises. IT security, ergonomics and battery life also need to be further improved, and cost-effective, compact and lightweight optical solutions for overlaying the virtual and real worlds need to be developed. Last but not least, the topic of standardization and standardization of augmented reality and data glasses must be addressed.

When will the results achieved in the Glass@Service project find their way into commercial products?

Schiefelbein : Within the current funding landscape, there are no budgets available to bring commercial products to market. Microsoft, for example, has invested around 150 million euros in HoloLens 1 . However, each partner will incorporate the results of the project into their own developments, products and processes. For example, the optical simulations carried out by Dioptic can be transferred to other applications within the service and product range. Uvex will also incorporate the project results into its digital protective equipment. And we at Siemens will use the knowledge gained from the use cases to increase productivity and improve quality in real logistics and manufacturing processes.