@inproceedings{Grohmann22:JAVRIS,

title = {JAVRIS: Joint Artificial Visual Prediction and Control for Remote-(Robot) Interaction Systems},

author = {Andreas I. {Grohmann} and Johannes V. S. {Busch} and Adrian {Bretschneider} and Christopher {Lehmann} and Frank H. P. {Fitzek}},

year  = {2022},

date = {2022-01-08},

urldate = {2022-01-08},

booktitle = {2022 IEEE 19th Annual Consumer Communications & Networking Conference 

(CCNC) (CCNC 2022)},

address = {Las Vegas, USA},

abstract = {In recent years, robots are taking on more and more tasks throughout many different application domains. Many of those robots work fully automatically and without human intervention. However, more complex tasks still require to be done by a human in remote control. One example would be robot remote surgery. Remote controlling a robot requires ultra-low latency on the communication, to allow fast and precise movements. To make this possible with today's systems, the human operator must be in the same room. Enabling the operator to operate from anywhere around the world, would bring a huge benefit as travels for highly trained experts can be minimized. In order to allow wider distances between the operator and robot, we propose an AI-based prediction. Predicting the robots behavior can generate negative latency, which improves the precision of control. In large-scale communication networks, the main part of experienced latency comes from the propagation delay and is therefore not avoidable. Predicting upcoming data before it actually arrives, can create a zero-latency experience for the human operator. To prove this idea in the context of remote robot control, we propose JAVRIS. Using the CMIYC demonstrator as example, JAVRIS improved the score of an inexperienced user by over 1000%.},

keywords = {},

pubstate = {published},

tppubtype = {inproceedings}

}