How long does it take to load the videos in On-Road IoT?
AI events: Upload is triggered immediately. Availability in On-Road IoT is dependent on mobile coverage and connection speed.
OBC events: Between 5 and 15 minutes after trip end. The VisionAI unit is configured with a delay to stay on for 15 minutes after ignition is switched off to allow for the download of videos.
How long are the videos available in On-Road IoT?
Videos are retained for a default period of 8 months. Once the video retention period expires, the videos will be deleted from S3 using an existing AWS process.
What is the size of the video files?
A single event of 8 seconds for 2 cameras is about 1.2MB for video footage using the standard image quality.
Where is it getting the video file from?
Video files are stored on our Amazon Web Services infrastructure after they are retrieved from the in-vehicle unit.
Does the VisionAI unit need the ignition on to upload video files to the MiX cloud infrastructure?
Yes.
Must the vehicle be online and communicating?
The VisionAI unit needs to be ON for video files to be uploaded to the MiX cloud. For video on-demand requests, if the user requests video and the device is offline, the request will go in a queue and retry later.
The first generation VisionAI recorded footage and allowed supervisors to review any footage within the last 72 hours or more depending on the memory capacity. Will this feature still be available through On-Road IoT for VisionAI?
Yes, it is available and is still dependent on the memory card capacity. Note the following times in hours are approximations only and refer to “ignition-on” time which includes the 15-minute post-trip for video file transfer.
How long is the video footage for each event? Will it be the same as generation 1 VisionAI?
Event-based video clips are 8 seconds in duration, video-on-demand is 2 minutes 30 seconds and Hi-Res requests (when released) will be 15 seconds. These durations are the same as Gen 1 VisionAI.
Will the footage requested be in HD or is there a possibility to choose resolution output?
All video files for events and video-on-demand are in a fixed quality and resolution which is not HD (which we refer to as Q3). The HD or Hi-Res video request feature will be available in a future release.
Video-on-demand: will road-facing and driver-facing views be available side-by-side?
No. The video-on-demand feature is a request to download the videos – it triggers an email that is sent with links to the corresponding video files, clicking on the link will download the video to your computer and does not open the video in On-Road IoT.
Defining Hi-Res
- Road-facing Camera = 1080P @ 20 frames a second
- In-cab Camera= 920P @ 10 frames per second
- Driver Camera = 920P @ 20 frames a second
How many seconds of video footage is uploaded, before and after an event?
Video footage for AI and OBC events is 8 seconds in duration, 4 seconds before and 4 seconds after the event timestamp.
Video-on-demand footage is 2imutes 30 seconds in duration from the start time the user specifies. This duration applies to standard or hi-res video-on-demand requests.
Hi-Res requests for events, when released, will be 15 seconds in duration, 7 seconds before and 8 seconds after the event timestamp.
Trip Timeline and Video-on-Demand: Can I request Hi-Res video footage?
Yes, when the Hi-Res request feature is available. Note that the standard quality setting is the default for video-on-demand requests but the user can override that each time they make the request.