Actuator quality in the Internet of Things

In this paper we look at a generally unexplored area, the operational boundaries that determine perceived quality in actuators connected to the Internet of Things (IoT). Research in the area of IoT has placed emphasis on networked devices with sensing capabilities generally overlooking other types of applications. We believe that as the IoT grows it is important to understand what impacts the perceived quality of experience when a user commands networked actuators. Our research finds direct applicability in the control of cyberphysical systems such as those present in networked unmanned vehicles, autonomous control of automobiles or realtime industrial process control, among other options. Using a real packet-based actuating testbed along with qualitative and quantitative experimental data we study the quality experienced by actual users. We quantify how different network metrics affect perceived quality and propose the first characterization framework in the area.