Abstract

The dream of augmented reality is one where the world we see is enhanced: overlaid with virtual objects that are visually stimulating, fun to play with, or provide utility. The most accessible parts of that dream to date are face filters, games, and some apps and gadgets that try to sell furniture or clothes. Seeing how the "fun" part is getting plenty of attention, in this thesis I explore a practical use case. Smartphones that support AR can by definition track changes in their spatial position and orientation, which in turn can be used to control a robot by having it mirror those movements. This paper presents how I built a proof of concept, and evaluates the feasibility of the general idea. I designed a web application for smartphones, and use AR technology with Bluetooth to turn the phone into a wireless controller. I have taken care that the robot doesn't execute invalid or unsafe orders. I found that the chain of tools I propose is stable and reliable for most use cases, and the low cost of required hardware makes it easy for anyone to come up with future iterations. The round-trip-time from input to feedback, however, leaves some untapped potential to be gained if AR controlled robots were to challenge legacy handheld controllers.

Results/Implementation/Project Description

Conclusion

[ PDF (optional) ] 

[ Slides Kickoff/Final (optional)]