Augmented Reality

Augmented Reality (AR) is a technology to display information in form of holograms directly into a person's field of view. We develop and test novel AR apps for context-aware support and real-time user feedback.

Context-​aware Support

Procedural tasks are common in many professions and require precise execution. Augmented reality (AR) has shown promise in effectively assisting operators with such tasks. However, most AR systems only provide static information to the operator without considering their current situation or level of experience. We develop context-​aware AR support to improve support especially for complex operations. By comparing user behavior inferred by real-​time analysis of sensor data (e.g. eye tracking) with a process model, the current context and required support can be inferred to adapt displayed information and to provide real-​time feedback.

external pageRead more            external pageWatch video    

pdz_AR_Glasses
Context-aware support of operators by augmented reality glasses (e.g. Microsoft HoloLens 2) in complex maintenance (left) and surgical (right) procedures. (ETHZ_pdz)

Visualization Strategies

Using accurate and easy-​to-use visualizations for surgical navigation is important to assist surgeons effectively during a procedure. As part of the Hochschulmedizin Zürich (HMZ) Flagship Project SURGENT, we compared 5 different augmented reality (AR) visualizations for surgical drilling with variations in abstraction level (abstract or anatomic), position (overlay or small offset), and dimensionality (2D or 3D). We tested these visualizations in a study with 4 expert surgeons and 10 novices (residents in orthopedic surgery) on lumbar spine models covered by Plasticine.

external pageRead more

pdz_AR_Visualizations
Five different AR visualizations to support a surgeon in finding the accurate trajectory for drilling into a spine bone ranging from 3D anatomic to 2D anatomic to abstract visualizations. (ETHZ_pdz)

Hand Action Prediction

Human error in industrial and clinical applications can be associated with high risk and cost. For a simplified use case, we developed an AR app that can predict erroneous hand actions before they are executed in a simplified use case. By utilizing the eye and hand tracking capabilities of the Microsoft HoloLens 2, our solution is able to anticipate what a user will do next and, in case of a mistake, warn the user in advance.

external pageRead more            external pageWatch video

pdz_AR_action_prediction
Implementation of an AR app for Microsoft HoloLens 2 using eye tracking (left) and hand tracking (right) to predict which card of a memory game is intended to turn next and to provide feedback before the turnover. (ETHZ_pdz)

Contact

Dr. Julian Wolf
  • PFA L 62

Chair of Product Dev.& Eng. Design
Technoparkstrasse 1
8005 Zürich
Switzerland

JavaScript has been disabled in your browser