Occupancy prediction using historical occupancy patterns

A system comprising:

• at least one sensor arranged to detect occupancy of a space;
• at least one memory arranged to store data collected by the at least one sensor;
• and
• at least one processor arranged to perform steps comprising:
• creating a table of past occupancy based on the sensor data;
• identifying a sequence of occupancy data for the space for an elapsed portion of a current calendar day and a portion of a day immediately prior to the current calendar day divided into discrete time slots;
• comparing the identified sequence to portions of sequences of occupancy data extracted from the table of past occupancy for the space to select one or more similar sequences, the portions of sequences of occupancy data extracted from the table of past occupancy for the space being selected using a progressive algorithm, each sequence of occupancy data extracted from the table of past occupancy relating to a calendar day and a portion of a previous calendar day divided into discrete time slots, use of the progressive algorithm comprising conducting a first comparison at a first level of granularity to identify a set of similar sequences and then conducting at least a second comparison against the set of similar sequences at a second level of granularity, the second level of granularity being a shorter time increment than the first level of granularity;
• computing a future occupancy probability for the space based on the selected one or more similar sequences;
• and
• sending a control message to operate a system based on the future occupancy probability for the space.

One or more device-readable media, the device-readable media being hardware, with device-executable instructions that, when executed by a computing system, direct the computing system to perform steps comprising:

• creating a table of past occupancy based on sensor data;
• identifying a sequence of occupancy data for a space for an elapsed portion of a current calendar day and a portion of a day immediately prior to the current calendar day divided into discrete time slots;
• comparing the identified sequence to portions of sequences of occupancy data extracted from the table of past occupancy for the space to select one or more similar sequences, the portions of sequences of occupancy data extracted from the table of past occupancy for the space being selected using a progressive algorithm, each sequence of occupancy data extracted from the table of past occupancy relating to a calendar day and a portion of a previous calendar day divided into discrete time slots, use of the progressive algorithm comprising conducting a first comparison at a first level of granularity to identify a set of similar sequences and then conducting at least a second comparison against the set of similar sequences at a second level of granularity, the second level of granularity being a shorter time increment than the first level of granularity;
• computing a future occupancy probability for the space based on the selected one or more similar sequences;
• and
• sending a control message to operate a system based on the future occupancy probability for the space.