Community

Hello, I noticed that the "Get all camera media" endpoint always returns the exact same image URLs for frontFacing and for driverFacing respectively within a response, even though the captureTime timestamps change. The URLs are identical. Modifying the endTime (not the startTime) may change the image returned. I think there's something wrong on your end. Potentially, the image corresponding to the acquired image closest to the endTime parameter is the one returned for all timestamp within the startTime-endTime timespan. Sample: ```javascript { "captureTime": "2021-02-22T17:18:04.187Z", "url": "https://s3.eu.samsara.com/samsara-eu-dashcam-videos/redacted } { "captureTime": "2021-02-22T17:23:05.186Z", "url": "https://s3.eu.samsara.com/samsara-eu-dashcam-videos/redacted } { "captureTime": "2021-02-22T17:28:05.187Z", "url": "https://s3.eu.samsara.com/samsara-eu-dashcam-videos/redacted } { "captureTime": "2021-02-22T17:33:05.187Z", "url": "https://s3.eu.samsara.com/samsara-eu-dashcam-videos/redacted } { "captureTime": "2021-02-22T17:33:59.187Z", "url": "https://s3.eu.samsara.com/samsara-eu-dashcam-videos/redacted } ``` This output is obtained from the following command line: ```shell curl --location --request GET 'https://api.eu.samsara.com/fleet/vehicles/cameras/media/history?startTime=2021-02-22T00:00:00.000Z&endTime=2021-02-22T23:59:59.999Z' \ --header 'Authorization: Bearer REDACTED' | jq -r '.data[0].cameras[0].images[] | {captureTime: .captureTime, url: .imageData[] | select(.cameraView == "frontFacing") | .url}' ``` Note: all redacted URLs are identical, which is the issue explained above. Changing the endTime may yield another image ```shell curl --location --request GET 'https://api.eu.samsara.com/fleet/vehicles/cameras/media/history?startTime=2021-02-22T00:00:00.000Z&endTime=2021-02-22T13:59:59.999Z' \ --header 'Authorization: Bearer REDACTED' | jq -r '.data[0].cameras[0].images[] | {captureTime: .captureTime, url: .imageData[] | select(.cameraView == "frontFacing") | .url}' ``` Additional question: is there a way to request the immediate capture of a new snapshot from the dashcam (front and driver)? This is very important for our use case. If not, what is the capture frequency? Looking at the captureTime timestamps, it's irregular: 5 minutes seems to be the baseline frequency but sometimes it's more frequent. Can snapshots be automatically be taken when engine is turned off? Thank you

Hi I'm currently trying to use https://api.samsara.com/fleet/vehicles/cameras/media/history but having issues with startTime and endTime parameters I set it as close as I can but it says missing from query string. python 3.x camera = 'https://api.samsara.com/fleet/vehicles/cameras/media/history' headers = &#123; "Accept": "application/json", "Authorization": "Bearer <<token>>", "startTime": "2021-02-14T00:00:00Z", "endTime": "2021-02-15T00:00:00Z" &#125; response = requests.request("GET", camera, headers=headers) print(response.text) result: <class 'str'> 2021-02-16T17:05:09Z 2021-02-16T17:05:09Z &#123;"message":"\"startTime\" is missing from query string; \"endTime\" is missing from query string","requestId":"74dcdd32-6e4ae656"&#125;

Can video content from Samsara cameras be diverted (streamed in real-time) to a third party onboard processor for interpretation? The use case I am looking at is a POC that would run in parallel to the verticalized Samsara configuration (camera-gateway-cloud). As a follow-up question, is Samsara considering allowing third-party AI/algorithms to run on their in-vehicle (vs. cloud) tech stack?