A Methodology for Cloud Robotics Implementation Based on Open-Source Software: from Human-Robot Interaction to Autonomous Applications
Nombre: RICARDO CARMINATI DE MELLO
Fecha de publicación: 26/11/2020
Supervisor:
Nombre |
Papel |
|---|---|
| ANSELMO FRIZERA NETO | Advisor * |
Junta de examinadores:
| Nombre |
Papel |
|---|---|
| ADRIANO ALMEIDA GONÇALVES SIQUEIRA | External Examiner * |
| ANSELMO FRIZERA NETO | Advisor * |
| CAMILO ARTURO RODRIGUEZ DIAZ | Internal Examiner * |
| CARLOS ANDRES CIFUENTES GARCIA | External Examiner * |
| EDUARDO ROCON DE LIMA | External Examiner * |
Páginas
Sumario: The advent of the cloud robotics paradigm has the potential to unleash a whole new generation of smart robotic devices by allowing robots to communicate with each other and with remote computing platforms to share experiences, sensor data, and to offload heavy processing applications. This work explores cloud robotics by casting a light on key issues and proposing novel solutions to allow for the widespread adoption of cloud-based functionality for research and commercial robots. We investigate the cloud robotics paradigm as an enabler to a series of applications and devices and question how the insertion of network and cloud technologies into such solutions might affect the interaction between a robot and a human operating it and what are the limiting requirements for cloud-based solutions. We propose a reference architecture for cloud robotics paired with an integrated network solution to allow for uninterrupted connectivity in mobile robotics applications. Furthermore, we present an open communication framework to link robots and cloud platforms, and that suits a large range of applications. Finally, we introduce a methodology for cloud robotics implementation based on open-source software and commercial off-the-shelf devices. Such a methodology leverages our communication framework and provides a common standard that allows reproducing and benchmarking different cloud robotics works. Our methodology and communication framework are validated considering different applications and multiple cloud configurations, showing the suitability of our approach and providing insight for other researchers and practitioners.
