SharpFlying: Design and implementation of a generic multi-service framework for autonomous indoor flight & An exploratory study into autonomous indoor Human-Drone Interaction
Translated title
SharpFlying
Authors
Helsted, Kasper Østergaard ; Carlsen, Steffen Darby
Term
4. term
Education
Publication year
2019
Submitted on
2019-06-11
Pages
16
Abstract
Forskningen i autonome droner, især til indendørs navigation, er vokset markant. Mange studier forsøger at løse opgaven med én enkelt teknologi, men få kombinerer flere tilgange i ét samlet system. Vi designede og implementerede SharpFlying, et generisk og udvidbart multi-service-rammeværk (et system, der integrerer flere funktioner) til indendørs, autonome droner. For at validere rammeværket udviklede vi tre proof-of-concept-tjenester og testede dem både hver for sig og i kombination. Resultaterne viser, at en kombination af flere tjenester giver markant bedre ydeevne end en enkelt løsning. Samtidig mangler der klare retningslinjer for, hvordan mennesker bør interagere med indendørs, autonome droner. Derfor gennemførte vi et tre-delt, eksplorativt studie af menneske–drone-interaktion med fokus på interaktion under navigation, stemmeinteraktion og sekundær interaktion. Deltagerne forventede, at dronen opførte sig pålideligt og tillidsvækkende; nogle beskrev ”tillid” som en metafor for, hvor fejlfrit de forventede, at dronen skulle agere. Vores studie skitserer en række designindsigter, der kan guide fremtidig udvikling af interaktion med indendørs, autonome droner.
Research on autonomous drones, especially for indoor navigation, has grown rapidly. Many projects pursue a single-technology solution, but few integrate multiple approaches into one system. We designed and implemented SharpFlying, a generic and extensible multi-service framework (a system that integrates several capabilities) for autonomous indoor drones. To validate the framework, we built three proof-of-concept services and tested them both individually and together. Our results show that combining multiple services delivers significantly better performance than relying on a single solution. Beyond autonomy, there are still no clear guidelines for how people should interact with indoor autonomous drones. We therefore conducted a three-part exploratory study of human–drone interaction focused on interaction during navigation, voice interaction, and secondary interaction. Participants expected reliable, trustworthy behavior; several used “trust” as a metaphor for how flawless they wanted the drone to be. Our study outlines design insights to guide future interaction design for autonomous indoor drones.
[This abstract was generated with the help of AI]
Keywords
Documents