Incident command loses the feed mid-operation
Cellular saturates as the public arrives; the command video and location feeds freeze precisely when coordination matters most.
Public safety & emergency response
When the network is part of the disaster, communications can’t be.
For civil-protection agencies, police, fire and EMS technology leads, first-responder communications teams and municipal resilience officers: incident communications that continue while public networks are congested, damaged or unevenly available across the response area.
The stakes
The stakes
The first hours of an incident are exactly when public networks perform worst.
The forward team is beyond coverage entirely
The operation doesn’t stop at the edge of the coverage map — but the communications plan usually does.
The first repaired tower is hours away
In a flood or storm zone, restoration is measured in hours to days. The response is now.
One connection over everything
Bond what remains. Mesh what you brought.
Nexus Atlas combines surviving public capacity — any carrier, any bearer — with the network the response brings: vehicle nodes, portable kits, satellite terminals. Field applications don’t change; they just stop caring which path is alive.
multi-carrier LTE/5Gportable satellitevehicle-to-vehicle meshprivate radiovenue Wi-Fi / wired
Three pillars
Built for degraded infrastructure.
Each pillar is a demonstrable behavior with the agency’s own workflows.
Continue over whatever remains
Saturation of one path is degradation, not an outage: traffic shifts gradually to the carriers and bearers still carrying, with command and location traffic first in line.
Bring the network with you
Vehicles and portable nodes form a self-healing mesh on scene — no infrastructure required, relaying up to 8 hops, reforming automatically as units move.
Nothing is lost in the gap
Reports, recordings and logs wait out a blackout in store-carry-forward — on disk if you choose, surviving a restart — and deliver automatically the moment any path returns.
0.1–0.75 sTypically configured reweighting cadence under congestion.
8 hopsRelay depth across vehicles and portable nodes.
6 classesDelivery contracts — command and location always first.
100% softwareRuns on vehicle routers and kit computers the agency already owns.
The command feed outlives the congestion.
Command, location and voice ride the priority classes across two carriers and a satellite terminal; public-network saturation slows the video, never the command channel. When the drone feed and a helmet camera share the video lane, the one set as primary is the last to lose a frame.
Reaching past the coverage edge.
A forward team relays through parked vehicles back to command — each vehicle an Atlas node, the chain reforming as vehicles reposition.
The mesh is the network for the first 12 hours.
With towers down, the response’s own nodes carry everything; when a carrier returns, it joins the bond as one more measured path — no cutover event.
A search drone over the flood plain sends its live picture over whatever links remain, and the picture drops late frames rather than stall. The drone keeps every packet of the stream; the command post asks for exactly the pieces it missed, and they come across on the lowest-priority lane while the drone flies or after it lands, so the footage the incident review works from has its gaps filled. How the gaps are filled
When the water takes a vehicle park, the operator draws the area on the map, reviews the units last seen inside it, and removes them in one step; a handset lost on scene that comes back online wipes itself on a signed command. Removal by area
A volunteer joins with one approval, and leaves with one removal.
A volunteer unit arrives with its own Atlas node and nobody has time to provision it. The node knocks; the operator sees who knocked, from where and how often, and approves it on the dashboard with the least it needs — the volunteer has a session about two seconds later, with no restart anywhere, and the decision is in the journal with the operator’s name. When the incident closes, one action removes the unit: every node ends its sessions with it and refuses it from then on, and the response’s shared secret is replaced. The approval queue How removal works
A partner agency watches one mission’s feed.
A neighbouring agency’s liaison officer is handed a link, not a radio. The incident’s drone feed leaves the response’s own network through a gateway that only sends outward, to a feed server the officer reaches in a browser; the player draws the officer’s name across the picture, the link works for the duration of the mission, and it is cut at the end with one click — or the moment it is no longer wanted. The response network opens no port, and nothing from the viewer’s side can reach the aircraft. How the gateway works
Fire’s network and police’s network, joined at the command post.
Each agency keeps its own network, keys and devices. The incident command post is a member of both and bridges them: police see the fire service’s drone feed at an address of their own network, and nobody reconfigures a single responder device. Each agency chooses which of its hosts take part, and either can withdraw the command post when the incident closes. Opt-in. How the seam works
Pure software. Pilot on the vehicle computers you already run — no procurement. EU-built and sovereign; the data plane needs no internet and no cloud to function on scene. Security & compliance.