Systems and methods of automatically estimating restaurant wait times using wearable devices

A computer-implemented method of monitoring activity at a location such as a restaurant, for estimating at least one or more different wait times at the location, comprising:

• identifying, by one or more computing devices, from at least one or more portions of sensor data from one or more mobile devices operated by a user that the user is seated at a restaurant table at an estimated sitting time, wherein the one or more portions of sensor data comprise motion sensor data, wherein the one or more mobile devices operated by the user comprise a wearable device, and wherein identifying that the user is seated at a restaurant table comprises identifying the occurrence of specific motion patterns from the one or more portions of motion sensor data, the specific motion patterns comprising one or more of the user placing his hands on the table, a reduced amount of hand motion of the user, or the user drinking;
• identifying, by the one or more computing devices, from at least time-correlated location data from the one or more mobile devices operated by the user an estimated arrival time of the user at the current location;
• wherein said one or more computing devices includes a machine learning classifier for defining a data model, the method further comprising:
• training the machine learning classifier to define a data model indicative of one or more event occurrences for the one or more different wait times, said training including:
• determining, by the one or more computing devices, an estimated first wait time by analyzing the difference between the estimated sitting time and the estimated arrival time;
• and
• determining, by the one or more computing devices, if one or more events has occurred to trigger the start of respective one or more wait times that are different than the first wait time, and if so estimating at least one different wait time based on the start time of respective one or more event occurrences;
• and
• applying the defined data model obtained from the machine learning classifier to automatically identify subsequent instances of said one or more event occurrences by providing the one or more portions of sensor data comprising the motion sensor data as input to the data defined model;
• and
• communicating, by the one or more computing devices, data indicative of the estimated first and at least one different wait time for the user at the current location to another computing device.