Abstract: A plurality of time and motion data sets are correlated at a network connected device. The time and motion data sets track an item of interest, and at least one of the time and motion data sets comprises energy consumption data or wireless local area network location data. Physical presence for the item of interest is determined based upon the correlating of the multiple time and motion data sets. Based upon the physical presence of the item of interest, an unacceptable condition for the item of interest may be determined. The unacceptable condition for the item of interest is remedied.

Abstract: A plurality of data sets are compiled in a memory. A first data set comprises a location mapping of an item of interest within a predetermined environment, and a second set comprises energy usage data within the predetermined environment. The first data set is correlated, via a processor, with the second data set to locate a subset of energy usage data in the second data set associated with the item of interest. A third data set is generated by combining the first data set with the subset of energy usage data in the second data set to increase an accuracy of the location mapping of the item of interest based upon the subset of energy usage data. The third data set is applied to a computing resource.

Abstract: Presented herein are techniques for managing a plurality of machine learning engines. A method includes receiving at an orchestration entity information descriptive of attributes of a plurality of machine learning engines, generating, based on the information, a unique signature for each machine learning engine of the plurality of machine learning engines, creating, based on the unique signature for each machine learning engine, an ensemble of machine learning engines configured to operate on a predetermined task, causing the ensemble of machine learning engines to operate on the predetermined task, and monitoring performance metrics of the machine learning engines in the ensemble of machine learning engines while the ensemble operates on the predetermined task.

Abstract: In one embodiment, a stream of data packets originated by a visual data source is received at an edge device in a network. The data packets include at least one of video data, image data, and geo spatial data. Next, a visual data attribute is extracted at the edge device from the stream of data packets according to an edge-based extraction algorithm. The extracted visual data attribute is vectorized at the edge device via quantization vectors. The vectorized visual data attribute is then indexed at the edge device in a schema-less database that stores indexed visual data attributes.

Abstract: A plurality of time and motion data sets are correlated at a network connected device. The time and motion data sets track an item of interest, and at least one of the time and motion data sets comprises energy consumption data or wireless local area network location data. Physical presence for the item of interest is determined based upon the correlating of the multiple time and motion data sets. Based upon the physical presence of the item of interest, an unacceptable condition for the item of interest may be determined. The unacceptable condition for the item of interest is remedied.

Abstract: A plurality of data sets are compiled in a memory. A first data set comprises a location mapping of an item of interest within a predetermined environment, and a second set comprises energy usage data within the predetermined environment. The first data set is correlated, via a processor, with the second data set to locate a subset of energy usage data in the second data set associated with the item of interest. A third data set is generated by combining the first data set with the subset of energy usage data in the second data set to increase an accuracy of the location mapping of the item of interest based upon the subset of energy usage data. The third data set is applied to a computing resource.

Abstract: A health status for a user is determined at a first network connected device based on health history stored at the first network connected device. Energy usage of a second network connected device utilized by the user is monitored by the first network connected device via messages sent to the first network connected device. A change in the health status of the user is predicted by the first network connected device based upon the messages sent to the first network connected device. Operation of at least one of the second network connected device or another network connected device is modified in response to predicting the change in the health status of the user.

Abstract: Techniques are presented herein for scheduling meeting rooms for participants by a server in communication with user devices. The server receives a meeting trigger signifying a necessity for scheduling of a meeting. The server then determines that a meeting room is needed for the meeting. After determining that a meeting room is needed, the server queries the plurality of meeting rooms located within a proximity of the first user for characteristics of the meeting rooms. The server then filters a subset of the plurality of meeting rooms based on a set of constraints for the meeting. Any rooms whose characteristics do not meet the constraints are filtered out of the plurality of meeting rooms. The subset of the plurality of meeting rooms are then sorted based on their usage state. The server then reserves one of the meeting rooms from the subset for the meeting between the participants.

Abstract: In one embodiment, a stream of data packets originated by a visual data source is received at an edge device in a network. The data packets include at least one of video data, image data, and geo spatial data. Next, a visual data attribute is extracted at the edge device from the stream of data packets according to an edge-based extraction algorithm. The extracted visual data attribute is vectorized at the edge device via quantization vectors. The vectorized visual data attribute is then indexed at the edge device in a schema-less database that stores indexed visual data attributes.