	US 12,321,143 B1
	Premises automation user recommendations
Alex Nathan Capecelatro, Los Angeles, CA (US); Timothy Earl Gill, Denver, CO (US); Kevin Carper, Denver, CO (US); Edward John McKenna, Jr., Denver, CO (US); Derek Murphy, Highlands Ranch, CO (US); and Atticus Gifford, Glendale, CA (US)
Assigned to Josh.ai, Inc., Denver, CO (US)
Filed by Josh.ai Inc., Denver, CO (US)
Filed on Sep. 15, 2020, as Appl. No. 17/021,982.
Claims priority of provisional application 62/902,904, filed on Sep. 19, 2019.
Int. Cl. G05B 19/042 (2006.01); G06F 3/16 (2006.01); G06Q 50/163 (2024.01); G06N 20/00 (2019.01); G06Q 30/0601 (2023.01); G06Q 50/06 (2012.01)

CPC G05B 19/042 (2013.01) [G06F 3/167 (2013.01); G06Q 50/163 (2013.01); G05B 2219/2614 (2013.01); G05B 2219/2642 (2013.01); G06N 20/00 (2019.01); G06Q 30/0631 (2013.01); G06Q 50/06 (2013.01)]

32 Claims

1. A system, comprising:

a user display;

a memory or other hardware storage device configured to store for each of a plurality of subsystems comprising a premises automation solution a corresponding current controlled state information; and

a processor coupled to the user display and the memory or other hardware storage device and configured to:

select and provide as output an initial recommendation in an event that a set of historical context user actions is empty;

track a historical context user action, at least in part by segregating the historical context user action into a set of bins based at least in part on a historical time for a historical user action and a historical location for the historical user action;

wherein each bin in the set of bins is bucketed by time of day, day of week, and location;

wherein each time of day bucket in the set of bins includes a start time of day and an end time of day based at least in part on how limited historical time of day data is;

wherein each day of week bucket in the set of bins includes a start day of week and an end day of week based at least in part on how limited historical day of week data is;

wherein each location bucket in the set of bins includes a geolocation range considered to be a close location for the location bucket;

wherein the historical context user action comprises a historical weather condition during the historical user action;

record the historical context user action in the memory or other hardware storage device;

track a current context, at least in part by segregating the corresponding current controlled state information in the set of bins;

wherein the current context comprises a current weather condition;

match the current context to the set of historical context user actions, wherein the matching is based at least in part on a weighted score based at least in part on what actions were performed at a same time of day bucket, a same day of week bucket, the close location using the set of bins, and a similar current weather condition to the historical weather condition;

modify the initial recommendation based at least in part on the set of historical context user actions;

generate a current recommendation statement for a given user associated with a user identifier to perform a recommended operation with respect to one or more of said subsystems of the premises automation solution, based at least in part on the matching;

render as output to the user display: the current recommendation statement; and an UI (user interface) element for a current recommendation response, wherein the UI element includes at least one of the following: a voice toggle button, a voice control button, a simplified direct control view, a button, a follow-up response option, an ellipsis, a slider, and a dialog;

receive as input associated with the rendered UI elements for the current recommendation response a user input associated with the current recommendation; and

provide a command to a control output device configured to control a connected device in a home, wherein the control output device is at least one of the following: a smart home hub, a smart home device with direct control, and a sensor.