2.1 KiB
Cloud
Once Kerberos.io is installed on the Raspberry Pi, the machinery and webinterface are available. A user can browse to the IP address (e.g. http://192.168.0.13) of the Raspberry Pi to configure the machinery and view the events that were taken by the machinery. However to be able to browse to the Raspberry Pi, a user has to be in the same internal network as where the Raspberry Pi is located; otherwise the webpage wouldn't be accessible. To make the Raspberry Pi accessible for the public, you will need to do some additional configurations on your router.
Another solution would be to use the cloud service that Kerberos.io is offering; A user pays a monthly fee to use the Kerberos.io cloud service. You can attach multiple Kerberos.io instances to the cloud service with only one license.
Video
How it works?
Kerberos.io is using AWS S3 to store their images. When the machinery detected something, it will save an image to disk, and maybe trigger some other IO devices. In its turn the machinery will push the image(s) to the appropriate S3 bucket, by using your credentials; the credentials you've received from the cloud application.
When the file (image) is stored in the bucket, AWS Lambda will do some validation and post-processing. Lambda will check if the file is realy an image, if the size is not to big, and do some post-processing like facerecognition. If the image is valid, the image will be copied to our production bucket, and additional information is stored in DynamoDB.
The cloud application is attached to the production bucket, so only images that are verified will be shown in the cloud application. When a user accesses the cloud application, the images from the production bucket are listed. Images that are loaded in the browser will be signed, for a specific time, before they are actually rendered. The signing will expire after 100 seconds.