Arrival detection for battery-powered optical sensors

1. A method comprising:

accessing an occupancy status of a workspace within a work area;

in response to the occupancy status indicating vacancy of the workspace:

at a first time, recording a first image of the workspace at a sensor block;

at a second time succeeding the first time by a first time interval, recording a second image of the workspace at the sensor block; and

detecting human arrival at the workspace based on features detected in the first image and the second image;

in response to detecting human arrival at the workspace:

at a third time, recording a third image of the workspace at the sensor block; and

detecting occupancy in the workspace, at the third time, based on features detected in the third image; and

in response to detecting occupancy in the workspace:

updating the occupancy status of the workspace to indicate occupancy;

at a fourth time succeeding the third time by a second time interval longer than the first time interval, recording a fourth image of the workspace at the sensor block; and

detecting continued occupancy in the workspace from the third time to the fourth time based on features detected in the fourth image.

2. The method of claim 1 , further comprising:

detecting vacancy in the workspace based on features detected in the fourth image;

at a fifth time, recording a fifth image of the workspace at the sensor block;

at a sixth time succeeding the fifth time by the first time interval, recording a sixth image of the workspace at the sensor block; and

detecting human arrival at the workspace based on features detected in the fifth image and the sixth image.

3. The method of claim 1 :

wherein accessing the occupancy status of the workspace comprises accessing the occupancy status of the workspace, in local memory of the sensor block;

wherein updating the occupancy status of the workspace comprises updating the occupancy status of the workspace indicating occupancy in the local memory of the sensor block; and

further comprising transmitting the occupancy status from the sensor block to a remote scheduling system.

4. A method comprising:

accessing a first occupancy status of a first workspace in a work area encompassing the first workspace and a second workspace;

during a first time period:

in response to the first occupancy status indicating vacancy of the first workspace, recording a first series of images of the first workspace at a sensor block and at a first imaging frequency;

during a second time period succeeding the first time period:

in response to detecting human arrival at the first workspace based on the first series of images, recording a second series of images of the first workspace at the sensor block and at a second imaging frequency less than the first imaging frequency;

in response to detecting occupancy of the first workspace based on the second series of images, updating the first occupancy status to indicate occupancy of the first workspace;

accessing a second occupancy status of the second workspace;

during a third time period succeeding the second time period:

in response to the second occupancy status indicating vacancy of the second workspace, recording a third series of images of the second workspace at the sensor block and at the first imaging frequency; and

during a fourth time period succeeding the third time period:

in response to detecting human arrival at the second workspace based on the third series of images, recording a fourth series of images of the second workspace at the sensor block and at the second imaging frequency; and

in response to detecting vacancy of the second workspace based on the fourth series of images:

maintaining the second occupancy status to indicate vacancy of the second workspace; and

recording a fifth series of images of the second workspace at the sensor block and at the first imaging frequency.

5. The method of claim 4 :

wherein recording the first series of images comprises, during the first time period, recording the first series of images of the first workspace at the sensor block and at the first imaging frequency, the first imaging frequency based on a first time of day of the first time period; and

wherein recording the second series of images comprises, during the second time period succeeding the first time period, recording the second series of images of the first workspace at the sensor block and at the second imaging frequency, the second imaging frequency based on a second time of day of the second time period.

6. The method of claim 4 :

wherein accessing the first occupancy status of the first workspace comprises accessing a set of occupancy statuses of a set of workspaces located in the work area, the set of workspaces comprising the first workspace;

further comprising, during the second time period, calculating a first total occupancy of the first workspace based on the first series of images;

wherein recording the second series of images comprises recording the second series of images of the first workspace at the sensor block and at the second imaging frequency, the second imaging frequency based on the first total occupancy of the first workspace;

further comprising, during the third time period, calculating a second total occupancy of the second workspace based on the second series of images; and

wherein recording the third series of images comprises recording the third series of images of the second workspace at the sensor block and at the first imaging frequency, the first imaging frequency based on the second total occupancy of the second workspace.

7. The method of claim 6 , wherein calculating the second total occupancy of the second workspace comprises calculating the second total occupancy of the second workspace greater than the first total occupancy of the first workspace, based on the second series of images.

8. The method of claim 4 :

further comprising, during the first time period, retrieving a last image in the first series of images of the first workspace at the sensor block;

wherein detecting human arrival at the first workspace based on the first series of images comprises:

executing an arrival detection model at the workspace based on the last image in the first series of images; and

detecting human arrival at the first workspace based on the arrival detection model;

further comprising, during the second time period, retrieving a first image in the second series of images of the first workspace at the sensor block;

wherein detecting occupancy of the first workspace based on the second series of images comprises:

executing an occupancy detection model based on the first image in the second series of images of the first workspace; and

detecting occupancy of the first workspace based on the occupancy detection model;

wherein updating the first occupancy status of the first workspace to indicate occupancy comprises:

interpreting occupancy of the first workplace based on the first image in the second series of images and the occupancy detection model; and

updating the first occupancy status of the first workspace to indicate occupancy; and

further comprising transmitting the first occupancy status to a remote scheduling system.

9. The method of claim 8 :

wherein executing the arrival detection model comprises executing the arrival detection model based on the last image in the first series of images of the first workspace, the arrival detection model characterized by a first energy consumption; and

wherein executing the occupancy detection model comprises executing the occupancy detection model based on the first image in the second series of images of the first workspace, the occupancy detection model characterized by a second energy consumption greater than the first energy consumption.

10. The method of claim 8 :

wherein retrieving the last image in the first series of images comprises retrieving the last image in the first series of images of the first workspace at the sensor block, the last image characterized by a first resolution corresponding to the first imaging frequency; and

wherein retrieving the first image in the second series of images comprises retrieving the first image in the second series of images of the first workspace at the sensor block, the first image characterized by a second resolution greater than the first resolution and corresponding to the second imaging frequency.

11. The method of claim 8 :

further comprising, during the third time period, retrieving a secondary last image in the third series of images of the second workspace at the sensor block; and

wherein detecting human arrival at the second workspace based on the third series of images comprises:

executing the arrival detection model based on the secondary last image in the third series of images; and

detecting human arrival at the second workspace based on the arrival detection model.

12. The method of claim 4 :

further comprising, during the first time period, retrieving a last image in the first series of images of the first workspace at the sensor block;

wherein detecting human arrival at the first workspace based on the first series of images comprises:

executing an arrival detection model at the workspace based on the last image in the first series of images;

detecting human arrival at the first workspace based on the arrival detection model; and

calculating a confidence score of human arrival at the first workspace based on the arrival detection model;

further comprising, during the second time period, retrieving a first image in the second series of images of the first workspace at the sensor block; and

wherein detecting occupancy of the first workspace based on the second series of images comprises:

in response to the confidence score falling below a threshold confidence score, executing an occupancy detection model based on the first image in the second series of images of the first workspace; and

detecting occupancy of the first workspace based on the occupancy detection model.

13. The method of claim 4 :

further comprising during the second time period:

retrieving a first image in the second series of images of the first workspace at the sensor block;

wherein detecting occupancy of the first workspace based on the second series of images comprises:

executing an occupancy detection model based on the first image in the second series of images of the first workspace; and

detecting presence of a human in the first workspace based on the occupancy detection model;

wherein updating the first occupancy status of the first workspace to indicate occupancy comprises, in response to detecting presence of the human in the first workspace based on the occupancy model:

interpreting occupancy of the first workplace based on the first image in the second series of images and the occupancy detection model; and

updating the first occupancy status of the first workspace to indicate occupancy; and

further comprising transmitting the first occupancy status to a remote scheduling system.

14. The method of claim 4 :

further comprising accessing a schedule defining:

a first time interval for recording the first series of images at the first imaging frequency;

a second time interval for recording the second series of images at the second imaging frequency; and

a third time interval for recording the third series of images at the first imaging frequency; and

wherein recording the first series of images of the first workspace at the sensor block and at the first imaging frequency comprises, in response to the first time period intersecting the first time interval of the schedule, recording the first series of images of the first workspace at the sensor block and at the first imaging frequency.

15. The method of claim 14 :

wherein recording the second series of the first workspace comprises, in response to the second time period intersecting the second time interval of the schedule, recording the second series of images of the first workspace at the sensor block and at the second imaging frequency; and

wherein recording the third series of images of the second workspace comprises, in response to the second time period intersecting the third time interval of the schedule, recording the third series of images of the second workspace at the sensor block and at the first imaging frequency.

16. The method of claim 4 :

wherein recording the first series of images of the first workspace comprises recording the first series of images of the first workspace at the sensor block via a first optical sensor;

wherein recording the second series of images of the first workspace comprises recording the second series of images of the first workspace at the sensor block via the first optical sensor; and

wherein recording the third series of images of the second workspace comprises recording the third series of images of the second workspace at the sensor block via a second optical sensor.

17. A method for detecting occupancy of a workspace comprising:

accessing an occupancy status of a workspace;

during a first time period:

in response to the occupancy status indicating vacancy of the workspace, recording a first set of images of the workspace at a sensor block, the first set of images characterized by a first imaging frequency and a first resolution;

during a second time period:

in response to detecting human arrival at the workspace based on differences in objects depicted in the first set of images, recording a second set of images of the workspace at the sensor block, the second set of images characterized by a second imaging frequency less than the first imaging frequency and a second resolution greater than the first resolution; and

in response to detecting occupancy of the workspace based on the second set of images:

updating the occupancy status of the workspace to indicate occupancy; and

transmitting the occupancy status to a remote scheduling system.

18. The method of claim 17 :

wherein accessing the occupancy status of the workspace comprises accessing the occupancy status of the workspace, in local memory of the sensor block;

wherein detecting human arrival at the workspace comprises detecting human arrival at the workspace based on differences in objects depicted in the first set of images at the sensor block;

wherein detecting occupancy of the workspace comprises detecting occupancy of the workspace based on the second set of images at the sensor block;

wherein updating the occupancy status of the workspace comprises updating the occupancy status of the workspace to indicate occupancy in the local memory of the sensor block; and

wherein transmitting the occupancy status to the remote scheduling system comprises transmitting the occupancy status from the sensor block to the remote scheduling system.

19. The method of claim 17 , wherein detecting human arrival based on differences in objects depicted in the first set of images comprises:

detecting a presence of a first contiguous edge in a first image in the first set of images;

in response to detecting an absence of the first contiguous edge in a second image in the first set of images, detecting human arrival at the workspace; and

recording the second set of images of the workspace at the sensor block.

20. The method of claim 17 :

further comprising accessing a template image of the workspace; and

wherein detecting human arrival based on differences in objects depicted in the first set of images comprises:

generating a first difference image between the template image and a first image in the first set of images;

detecting a human arrival in the workspace based on the first difference image;

generating a second difference image between the template image and a second image in the first set of images;

confirming the human arrival in the workspace based on the second difference image; and

recording the second set of images of the workspace at the sensor block.