US 12,316,502 B2
Creating sets of heterogeneous internet-of-things (IoT) devices from which to upload data as a pre-step to edge computing
Hemant Kumar Sivaswamy, Pune (IN); Venkata Vara Prasad Karri, Visakhapatnam (IN); Sarbajit K. Rakshit, Kolkata (IN); and Afroz Khan, I, Davanagere (IN)
Assigned to International Business Machines Corporation, Armonk, NY (US)
Filed by INTERNATIONAL BUSINESS MACHINES CORPORATION, Armonk, NY (US)
Filed on Nov. 30, 2022, as Appl. No. 18/060,197.
Prior Publication US 2024/0179062 A1, May 30, 2024
Int. Cl. H04L 41/0893 (2022.01); G16Y 40/30 (2020.01); H04L 41/0823 (2022.01)
CPC H04L 41/0893 (2013.01) [G16Y 40/30 (2020.01); H04L 41/0823 (2013.01)] 17 Claims
OG exemplary drawing
 
1. A computer-implemented method, the method comprising:
identifying one or more Internet-of-Things (IoT) devices within a structure;
combining the identified one or more IoT devices into one or more groups based on a proximity of their respective installation locations within the structure;
identifying a respective primary device within each group of the one or more groups;
deploying a tiny machine learning (ML) model on each identified respective primary device;
in response to detection of an event within the structure by a group of the one or more groups, utilizing the tiny ML model of a primary device of the group to select one or more other groups for activation, wherein selection of the one or more other groups is based, in part, on a type of the event, and wherein a selected other group detects events of the type;
transmitting, to a cloud resource, data of the event collected by the group;
broadcasting a streaming request to the selected one or more other groups to transmit, to the cloud resource, data of the event collected by the selected one or more other groups; and
broadcasting a no-streaming request to unselected groups of the one or more groups to restrict data transmission from the unselected groups.