Steinbeis experts redevelop indoor positioning and augmented reality (AR) applications
For Professor Dr.-Ing. Frank Deinzer of the Steinbeis Transfer Center for New Media and Data Science, the starting point is university research. Supported by his team, Deinzer is conducting research into artificial intelligence (AI) with a focus on sensor fusion. He tells TRANSFER magazine about his findings and how they are being applied to concrete AR projects.
Sensor fusion is the process of combining data from different – sometimes disparate – sources of information and sensors in order to derive an overall picture of a given situation. From a research perspective, this is a universal issue that leads to generally applicable methods. “You step closer to applications the moment the focus turns on specific, demanding problems and you’re able to draw on your research results,” explains Frank Deinzer.
Moving from the problem to the solution
For years, one such problem has been the issue of indoor positioning: Where is a person or object located inside a building? The items of technology people carry around anyway play an important part in this – smartphones, which perceive their surroundings through built-in sensor systems. If this information can be combined with what is already known – for example by using building layouts – this should be enough to continuously determine a person’s position, in sufficiently accurate detail.
Another topic the Steinbeis Enterprise is working on is augmented reality (AR), i.e. the interplay between analog and digital life. For example, there are applications that enrich image data from smartphone cameras with information generated artificially by the device. AI then helps identify and track objects or people in the image data stream. From a scientific standpoint, it is important to use sensor fusion techniques to do this.
Does this add value?
The question is: Ultimately, what does indoor positioning offer that genuinely adds value? One example would be location-based services, which have now become indispensable in outdoor areas. These are not just conventional navigation systems, but all of the kinds of applications offered by Google Maps – searching for the nearest store or services, location-based games and advertising, or recommended local events. Similar applications are virtually knocking down the door to indoor environments – navigation within large, complex buildings such as airports, audio guides in museums, or the analysis of pedestrian routes and highly frequented areas within buildings. These are precisely the applications being worked on at the Steinbeis Transfer Center for New Media and Data Science. Based on the findings of university research, the experts are setting up adaptable position-finding frameworks for use in specific projects.
So which environments should AR be used in for users to benefit from it, combined with which information? One example is a toolkit developed by the Steinbeis experts from Dettelbach to allow museums to create their own AR apps and add content without outside support. Unlike traditional guided tours, the new approach to evolving information allows visitors to explore the museum interactively, on a completely new level. “This also allows us to inspire younger museum visitors, making it possible to offer more exciting and contemporary educational content,” believes Steinbeis expert Toni Fetzer.
Successful knowledge and technology transfer
The broad range of potential uses for the applications is best demonstrated by looking at current projects. A navigation system is currently receiving its final touches at Würzburg City Hall, a complex extending over several wings and floors. Visitors can use a smartphone app to search for topics, departments, or people, which they are then guided to, right up to the actual door. The New Media and Data Science center also contributed to the 100th anniversary of the Mozart Festival in Würzburg, using the AR toolkit to create an app that allowed visitors to use their smartphones to enjoy an audiovisual experience of the exhibition, exhibits enriched with texts, audio and video streams, and historical content. Frank Deinzer and his colleagues also implemented a project for the Porsche Museum in Stuttgart, using an AR app to allow visitors to engage in an immersive exploration of a model measuring approximately seven by three meters at the Porsche Development Center.
AI slides into the background
As the three project examples show, AI does not exactly leap out at you when it is included in overall solutions. And that’s a good thing. The focus always lies in people and their requirements – everything else is secondary. It is precisely for this reason that the “best” AI is not even consciously noticed, even if it may be central to a key function of an app and really impresses users by its performance.
Would you like to learn more about the projects of the Steinbeis Transfer Center for New Media and Data Science? Take a look at the simpleLoc YouTube channel at https://simpleloc.de/yt 
or visit this website to try out the augmented reality app toolkit yourself: https://augmented-art.de