Introducing the FogSphere Manager
Artificial Intelligence is at the heart FogSphere Manager’s technology strategy. We seek to leverage the power of machine learning and computer vision, via both Edge/Fog video streams processing (within the customer local network boundaries) without requiring an internet connection to reach the cloud platform, to create powerful and flexible solutions for our clients.
Giving AI Power to existing network Cameras
FogSphere Manager is all about boosting your existing investment's value!
FogSphere makes your cameras smart with AI, without forcing you to renew your investment in replacing all your cameras, but we reuse as much as possible of the existing CCTV infrastructure.
Deploy at the Edge or in the Cloud
FogSphere® specialise in mobilising visual analytic technologies to enable them to be deployed in the most remote and challenging environments, by using the IoT management technology FogSphere® integrates remote sensors and actuators with the same hardware and software used to operate AI algorithms.
The FogSphere® family of solutions can be deployed on existing Edge, Fog or Cloud infrastructure.
We did AIoT - Native AI/IoT integration!
Here at FogSphere we've mastered and pioneered the seamless integration of AI and IoT in what we call an AIoT hybrid architecture. FogSphere enables triggering alarms and automations using events generated by our AI along with events generated by our IoT module (based on the LoRaWAN® technology).
The peace of mind of Working Offline
FogSphere® provides completely Disconnected Operations, allowing automatically alarms, triggers and generic workflows to run by the edge nodes directly, without requiring any cloud connection, hence limiting the incurrence of missed alarms cause by unstable connection to the cloud management platform.
One platform Many Architectures
The power of unlimited scale
On-Premise Fog deployments of the FogSphere Manager® are fully upgradable from single servers to full-scale, multi-server, clustered deployments thanks to our horizontally-scalable architecture based on Kubernetes and a software design based on micro-services and service-meshes.
FogSphere Manager® goes from single to multi-node installations simply by adding new ones to the original Kubernetes cluster:
-
- Data is automatically replicated over multiple cluster nodes
- Compute load is automatically balanced over the new cluster nodes joining the initial cluster, through the integrated load balancer
- Security is handled by WAF (Web Application Firewalls) and Network Firewalls provisioned along the FogSphere Manager® infrastructure
- AI Gateways do not experience any issue upon increase of cluster’s scale and load.
Fog installations can be entirely migrated to cloud installations at any moment. Edge AI gateways, Triggers & Alarms, Workflows, etc. are migrated from the local cluster to the fogsphere.com cloud organization deployed for the customer.
Built with micro-services
FogSphere has been built using the latest generation technologies for horizontally scalable micro-services. It is completely containerised and is operated with standard container orchestrators.
Scalable Storage
FogSphere relies on an internal storage abstraction that enables the horizontal scalability of both its internal Databases (for events, cameras, analytics) and Object Storages (for video recordings). Additional space and storage throughput can simply be provided by adding more nodes to any FogSphere cluster.
Unlimited AI Scale
As for the storage, also FogSphere's compute capabilities are not limited by hardware capabilities. Adding more Edge-AI devices (mainly Nvidia Xavier or Nvidia Orin inference accelerators) allows to linearly add more capabilities in terms of processing more cameras or more concurrent algorithms!