Augmented Reality Robot Operation Interface with Google Tango

In this paper we present a novel smart-device-based robot operation interface providing augmented reality (AR) and utilizing a device equipped with self-tracking functionality and a depth camera according to Google Tango reference design. The robot can be both programmed remotely by our Java application and taught by kinesthetic programming. The system also provides an object detection that is capable of recognizing objects. Additionally we designed and investigated an adaptive program execution. The ease of use is ensured by AR features such as the highlighting of detected objects and the preview of robot operations. The main contribution is a robot operation interface that can be multifariously programmed and is capable of adapting to changing surroundings based on commercially available devices.