Feathergrid at the Edge: Gateways, On-Site Decisions and Working Offline
Feathergrid collects sensor data at the edge through a gateway, can find what is related and line up the next move right there, and keeps working from the latest picture it has when the connection drops.
Feathergrid works at the edge. Sensor data is collected on-site through a gateway. Feathergrid can find what's related and line up the next move right there, without waiting on a round trip to the cloud. And if the connection drops, it keeps working from the latest picture it has.
Here's what that means, and where it matters.
What does "the edge" mean?
The edge is where the work actually happens: a field, a depot, a pump station, a kitchen, a loading dock. It's where the sensors are, and where someone has to act.
The usual pattern is to ship every reading to a cloud data center first, process it there, and send an answer back. That works until the connection is slow, patchy, or gone. Then the place that needs the answer is the last one to get it.
Feathergrid is built so the edge isn't left waiting.
How does data get from the field to Feathergrid?
Through a gateway on-site. Sensors connect to the gateway, and the gateway passes their readings to Feathergrid. The rest of what you run (your systems, apps and records) connects through Feathergrid's connectors, which are built to work with APIs.
Starting from nothing? The starter sensor pack, included with annual founding plans, ships a gateway with temperature, humidity and leak sensors already linked to your site. Plug it in and the readings show up in your picture.
Feathergrid can find what's related and line up the next move right there, without waiting on a round trip to the cloud.
What can run on-site?
Correlation and dispatch. Feathergrid can relate what the sensors are saying (same time, same place, one thing leading to the next, a pattern it has seen before) and line up the next move at the edge itself.
Your rules still decide what happens next. Routine steps run on their own. Big calls wait for a person. One switch stops everything.
What happens when the connection drops?
Feathergrid keeps working from the latest picture it has. A field crew, a remote site or a vehicle doesn't lose its picture just because the network did.
How does the edge fit with Azure and AWS?
The edge is one part of the same Feathergrid. The application and your data run on Microsoft Azure in the United States. AWS holds only the fingerprints that anchor the tamper-evident record, never your data. The edge follows the same rules as everything else, and every recommendation, approval and action belongs to the same record.
For the full picture of what runs where, see Feathergrid on Azure and AWS and the architecture page.
Where does the edge matter most?
Anywhere the work is far from a data center, or the connection can't be trusted:
- Agriculture. Soil, weather and field readings, and the call on when to treat or launch a drone. See agriculture.
- Field equipment. Pumps, compressors and machines that give early warning signs. See field equipment.
- Logistics. Trailers, docks and cold chain on the move. See logistics.
- Restaurants and food service. Coolers and kitchen equipment across many locations. See food service.
- Critical infrastructure. Utilities, water and emergency operations. See government and critical infrastructure.
The short version
Sensors talk to a gateway. Feathergrid can relate what they're saying and line up the next move on-site. Your rules decide what runs on its own. And when the connection drops, the work doesn't stop.
Topics
