Introduction of a Remote Lab for Indoor Object Position Control Based on Computer Vision Sensors and AI-Enabled Embedded Systems

Recent strides in the efficacy of Artificial Intelligence (AI), the adoption of software frameworks for AI and the power of Artificial Intelligence and Machine Learning (AIML) Embedded Platforms have come together to unlock the power of AI for embedded devices. The integration of these topics into degree program courses is mandatory for electronic engineers of the future. The University of Applied Science (UAS) Technikum Wien offers various bachelor and master degree programs dedicated to electronic engineering particularly with a focus on embedded systems software development based on compact AI inference embedded platforms. These degree programs are available in the form of full-time (accompanied with remote classrooms) and part-time. Moreover, the university promotes these degree programs to be adaptable for handicapped students as well. Due to the fact, that part-time and handicapped students do not attend the university on a regular basis and even full-time students have to attend remote classroom sessions, it is necessary that they have the opportunity to complete their hands-on exercises and assignments at home. These tasks often require the implementation of computer vision software modules based on various sensor technologies, utilization of (pre-)trained neural networks for embedded devices, implementation of embedded control algorithms and evaluation of their applications. To address this need of access outside the university, a suitable remote lab which provides convenient hardware-acceleration for AI-enabled embedded computer vision is necessary. This paper introduces a remote lab for indoor object position control based on various computer vision sensor technologies accessed via an AI-enabled embedded platform which has the potential to increase the availability of expensive AIML embedded platform hardware and closes the lack of access for students. Hence, this paper gives an overview of the actual indoor object position control application, the utilized sensor technologies and the graphical user interface.