You can configure your Samsara CM22 and CM32-series dash cams to use Camera ID to automatically assign drivers to vehicles using inward-facing dashcam footage and advanced machine learning. With Camera ID, Samsara automatically attributes an unassigned driver to a vehicle, trip, or harsh event for powerful driver-based insights.
Before the camera can automatically detect drivers, you must manually train the camera to identify drivers using driver images. Camera ID compares the video frames of the driver's facial features to the facial features of drivers that you manually assign. The system learns to assign the driver with features that most closely match the manually assigned driver's features; however, driver assignment is dependent on the driver hierarchy logic (see Driver Assignment Impact).
For best results and to configure your Camera ID driver associations, refer to the following topics:
Training the camera ensures accuracy when lighting, body pose, outfits, or other attributes change. Camera ID begins to auto-detect drivers after the first manual assignment. Camera ID takes time to learn who drivers are so you may not get an accurate driver suggestion for every segment. To ensure accurate detection:
-
Assign a minimum of 20 assignments.
-
Perform manual assignments and correct any inaccurate driver assignments.
-
Perform manual assignment for changes of driver appearance, such as differences to hairstyle, facial hair, hats, and glasses.
-
Train the algorithm to detect drivers wearing masks. Camera ID can still detect faces on drivers wearing masks with some reduced effectiveness.
After you configure Camera ID, the camera automatically associates the driver with the vehicle for vehicle and driver activities. Automatic assignment can take up to two minutes to populate in the Samsara dashboard. The Camera ID performs automatic assignment for the following events:
-
Any events associated with the vehicle for a trip, such as trip segments, the Trip History Report, the fuel and energy report, and the driver efficiency report
-
Safety events that surface in the Safety Inbox
Samsara uses hierarchy logic to determine which driver to assign to the vehicle. For example, an HOS driver assignment takes priority over a Camera ID driver assignment. For more information on assignment logic, see Driver to Vehicle Assignments.
Before you enable and configure Camera ID settings:
-
Review Camera ID Driver Consent. Laws in certain US states and EU member states may require you to get explicit written consent from your drivers before you enable this feature. For more information on facial recognition data management see Facial Recognition Privacy and Security .
-
Review the recommendations for mounting positions and verify the inward-facing camera is installed properly to reduce glare with a clear, unobstructed view of the driver.
-
Ensure that the driver list and driver profiles are current.
Before you enable Camera ID to automatically detect drivers, review the Camera ID Prerequisites.
-
Select the Settings icon (
) at the bottom of your Fleet menu to view dashboard settings.
-
From Devices, select Configuration > Cameras.
-
Enable the Camera ID setting.
-
Select the box and Confirm to acknowledge Camera ID consent.
-
Select the detection mode and determine method of driver assignment:
-
Full Fleet Mode: (Default) Use this mode for fleets of fewer than 100 drivers. When feasible, Samsara recommends that you perform initial Camera ID training with Manual Driver Confirmation Only to improve driver detection accuracy.
-
Grouped Mode: Best for large fleets of 100 drivers or more. Enables you to train the dash cam with smaller, more manageable groups using tags to limit driver assignment options. Requires Auto-Assignment Groups.
-
-
Save your changes.
-
To begin training the dash cam, assign dash cam segments to drivers (see Train Dash Cam Facial Recognition (Camera ID)).
The data for Camera ID is securely stored and protected in line with all of our other Samsara cloud data. Our cloud service is designed to meet the requirements of security-sensitive and privacy-sensitive organizations. Samsara does not disclose the personal information we process on behalf of customers, including Camera ID facial recognition information, outside of Samsara other than for the limited purposes described in our Terms of Service. Camera ID facial recognition information is used only to assign drivers to trips within their own organization and to improve our services.
Comments
0 comments
Please sign in to leave a comment.