Head of the Robotics and Machine Perception laboratory Roberts Kadiķis participates in the final event of the VIZTA project, where the results achieved in the three and a half year long project are presented. EDI’s task in the VIZTA project was to demonstrate the use of 3D/depth sensors created by the partners at an industrial task. The chosen task was sorting plastic bottles using a smart robot arm. Through research in computer vision, artificial intelligence and robot control algorithms, EDI was able to create a demonstrator consisting not only of an industrial robot made smarter and more capable during the project, but also of a mobile platform capable of detecting bottles on the floor, autonomously navigating to them and pushing the bottles outside the work area. EDI also researched synthetic data generation approaches for training artificial neural networks, as well as imitation learning techniques.