Dew-Computing in Future Telerobotic Applications: An Exploration

The inception of global pandemic has underscored and reinforced the utility and necessity of Telerobotic applications as an important factor to augment human endeavors in future. Such applications may broadly be classified into Telepresence or Teleoperation. Both function with human in the loop as an operator or observer. However, practical deployment experiences reveal that the expected advantage from such systems is largely unrealized due to the prevalent cloud-centric architecture. The key reason is the erratic response from the end-to-end channel over the Internet which is beyond the control of the end-applications. This calls for an exploration on how the promises of Dew-computing can be leveraged to ensure performance guarantee for such systems for the end-users agnostically. Particularly we shall explore how negotiation and process sharing with cloud for opportunistic localization of decision logics and storages may help compliment the desired QoS of the multi-sensory traffic-types being exchanged through formation of optimized topology among the robot, observer, and operator nodes. It will be also explored how Dew-based approach may help reduce session establishment overhead at a system-level. We shall explore two specific broad approaches, namely, Dew-on-Robot and Dew-on-Edge, and show how they fit into different dynamics of Telerobotic applications, especially in enterprise settings.