Augmented Reality (AR) enables the augmentation of the user's environment through visual provision of additional information. AR devices, such as glasses are able to use machine learning (ML) to grasp and analyze their environment and thereby interact with it. In addition to the consumer sector, AR technology has great potential, especially for industry and medicine.
In an industrial environment, AR systems can e.g. contribute to an individualized and flexible production and assembly by projecting assembly instructions onto the installer's AR glasses. Thus, the fitter is able to flexibly process every conceivable product or its configuration without the necessity of knowing very assembly plan in detail.
AR in medicine
In a medical context, AR systems can help to learn surgical procedures and medical interventions on models more precisely and more realistically (e.g. with step-by-step instructions). AR systems can also support doctors in the operation by covering the patient's body with real-time data from imaging devices such as CT, MRI, ultrasound or X-rays, thereby making organs and bones visible.
Several requirements must be met for the use of AR systems, applying equally to the industrial and medical context. It is important to keep immersion - i.e. the feeling within an AR system - extremely natural. An example of this is the subjective perception of a correct and natural visualization of the AR system. The visualization smoothly brings the real movement, environment and situational conditions into the system. This is also called context-conscious tracking and includes abstract positioning as well as the introduction of orientation information. In order to meet these requirements, the positioning must take demanding technical aspects into account.
The accuracy and precision of both localization and orientation are subject to strict limits. The accuracy and precision of the localization must be as high as possible. The more precisely the real and virtual objects are to match, the higher the accuracy and precision of the tracking algorithm must be.
At Fraunhofer IIS, research is not only carried out on high accuracy, but also on how a tracking system can be designed immersively and how localization and orientation can be kept robust and stable over a long period of time, especially in a non-linear and dynamic environment.
5G positioning for AR applications
In addition to communication, 5G also has the potential to offer added value for positioning in AR applications by gradually striving for higher positioning accuracy and latency from Release 17, which is suitable for supporting use in AR applications.