WorkEdge and cloud
IP Timelapse
Cloud-hybrid timelapse SaaS for distributed organizations: edge workers capture, the cloud renders.
- 0
- inbound ports opened on a site's network: edge workers dial out to the cloud core
- Org → site → edge
- every site, edge credential and stored frame belongs to an organization; an upload that doesn't match its edge's own site is refused and logged
- 10 s
- edge heartbeat with its camera list, capture-worker health and upload backlog; the cloud flags stale cameras, low disk and upload lag
- 172 of 172
- frames delivered after a storage outage test: the edge buffers in a durable queue, with no loss and no duplicates
IP Timelapse is a cloud-hybrid platform for organizations with cameras spread across many sites, deployed in production. At each site, a containerized edge worker on a small local machine connects to the IP cameras over ONVIF, pulls frames over RTSP on a fixed interval and uploads them to object storage. The cloud core runs the web interface, tracks every edge and camera, and uses ffmpeg to turn weeks of frames into a short video.
The data model is built for distributed organizations. Every site, edge credential and stored frame belongs to an organization, and the cloud refuses any upload that doesn’t match the sending edge’s own site. Enterprise role-based access control is designed on that foundation: roles such as owner, site admin, operator and viewer, with permissions down to a single feature, such as live view or download, scoped to an organization, a site, an edge or one camera.
Each edge reports its health every ten seconds: cameras, capture workers and the upload backlog. The cloud checks every camera for stale frames, watches disk space and upload lag, and shows the result on the dashboard. Edges only dial out, so nothing has to be opened on a site’s network. Camera passwords are encrypted at rest, and logins are rate-limited with lockout.
Video analytics is designed as the next layer, with a common event format and a pipeline from camera to cloud.