System and method for voice based controlling a tele-robot in a multi-user telepresence session

A processor implemented method, comprising:

• enabling a speech-command interface and establishing a dedicated secured bidirectional WebSocket P2P connection between the ASR cloud and one of the plurality of user—WebRTC clients identified as an operator client;
• receiving, by the ASR cloud, a first recorded speech signal for a pre-defined window size from the operator client, wherein if the first recorded speech signal contains a pre-defined wake-up word, the ASR cloud transmits a positive response signal to the operator client;
• recording, by the operator client, a second speech signal from the operator client, upon receipt of the positive response signal, wherein the second speech signal comprises of a control command for a tele-robot;
• transmitting the second speech signal to the ASR cloud, by the operator client;
• generating, by the ASR component in the ASR cloud, an inferred command from the second speech signal, wherein the inferred command is transmitted back to the operator client;
• obtaining, at the operator client, a validation response indicating whether the inferred command is aligned to the task to be performed;
• transmitting, by the operator client, the inferred command to the tele-robot—WebRTC client, through a dedicated secured bidirectional P2P WebRTC data channel between the tele-robot—WebRTC client and the operator client, if the validation response indicates that the inferred command is aligned to a task to be performed;
• triggering the edge device to generate a set of navigation commands from the inferred command, by the tele-robot—WebRTC client, wherein the generated set of navigation commands is transmitted to the tele-robot—WebRTC client;
• and
• performing, by the tele-robot—WebRTC client, navigation for the execution of the task, based on the generated set of navigation commands.

One or more non-transitory machine-readable information storage mediums comprising one or more instructions which when executed by one or more hardware processors cause a task execution, by:

• enabling a speech-command interface and establishing a dedicated secured bidirectional WebSocket P2P connection between an Automatic Speech Recognition (ASR) cloud, and one of a plurality of user—WebRTC clients identified as an operator client;
• receiving, by the ASR cloud, a first recorded speech signal for a pre-defined window size from the operator client, wherein if the first recorded speech signal contains a pre-defined wake-up word, the ASR cloud transmits a positive response signal to the operator client;
• recording, by the operator client, a second speech signal from the operator client, upon receipt of the positive response signal, wherein the second speech signal comprises of a control command for a tele-robot;
• transmitting the second speech signal to the ASR cloud, by the operator client;
• generating, by an ASR component in the ASR cloud, an inferred command from the second speech signal, wherein the inferred command is transmitted back to the operator client;
• obtaining, at the operator client, a validation response indicating whether the inferred command is aligned to the task to be executed;
• transmitting, by the operator client, the inferred command to a tele-robot—WebRTC client, through a dedicated secured bidirectional P2P WebRTC data channel between the tele-robot—WebRTC client and the operator client, if the validation response indicates that the inferred command is aligned to the task to be executed;
• triggering an edge device to generate a set of navigation commands from the inferred command, by the tele-robot—WebRTC client, wherein the generated set of navigation commands is transmitted to the tele-robot—WebRTC client;
• and
• performing, by the tele-robot—WebRTC client, navigation for the execution of the task, based on the generated set of navigation commands.