Skip to content
Skip to contentNexus Observatory
Nexus Observatory

/

Operations centre

/

Architecture
How operations centre is built.

One reference architecture under every deployment: the layers of a site, the protocol functions this platform relies on, and the evidence that travels with every observation.

Plan with usIndustries

Areas · 5 Technologies · 192 Industries · 3

At a glance

Confidential computing and hardware trust

In Operations centre, the confidential computing and hardware trust layer holds trusted execution environments, remote attestation and device identity and roots of trust, each drawn against the standard it follows. Hardware-isolated enclaves and confidential VMs that protect data in use from the host.

সমস্যা

Every node watched. Every incident closed

This page shows how Operations centre is put together, layer by layer, from the instrument to the record. Monitoring of fleets, sites and links around the clock where contracted, incident response from first alarm to root cause, managed services that run the network for you, and service levels written as numbers you can measure on your own dashboard.

Components

  • Pole-mounted edge gateway node
  • Particulate monitoring station and sensor node
  • Radar stream gauge on a road bridge
  • Street small-cell pole with an air-quality node
  • “AI tower in a box”
  • Data hall row pair with cold-aisle containment
  • Network operations centre with a curved video wall
  • Field-service crew vehicle

Interfaces

উপকরণসমূহ
SDI-12 · RS-485 / Modbus RTU · 4–20 mA
Radio
Ethernet · LTE-M / NB-IoT · LoRaWAN
Time and position
GNSS 1PPS
Security
TPM 2.0
Publishing
MQTT

Where it works

  • Servers and accelerators
  • Modular edge data centres and AI towers
  • OTA updates, GitOps and software supply chain

In the story below

  1. The system, part by part
  2. Assembly
  3. Installation
  4. Deployment
  5. Connections
  6. Operation
  7. Failure and recovery
  8. Scale and federation
  9. What you get

Scenes · 13

Next steps

Plan a deploymentOpen the Field Deployment StudioSee Operations centre
  1. Operations centre · Architecture

    Every node watched. Every incident closed

    This page shows how Operations centre is put together, layer by layer, from the instrument to the record. Monitoring of fleets, sites and links around the clock where contracted, incident response from first alarm to root cause, managed services that run the network for you, and service levels written as numbers you can measure on your own dashboard.

  2. Architecture · Compute and edge

    Confidential computing and hardware trust

    In Operations centre, the confidential computing and hardware trust layer holds trusted execution environments, remote attestation and device identity and roots of trust, each drawn against the standard it follows. Hardware-isolated enclaves and confidential VMs that protect data in use from the host.

  3. Architecture · Compute and edge

    Servers and accelerators

    In Operations centre, the servers and accelerators layer holds short-depth and ruggedised edge servers, GPUs and NPUs and inference accelerators, each drawn against the standard it follows. Servers sized for telecom cabinets with front access, wide temperature range and DC power.

  4. Architecture · Compute and edge

    Edge tiers and multi-access edge computing

    In Operations centre, the edge tiers and multi-access edge computing layer holds device edge, far edge and near edge, each drawn against the standard it follows. Compute inside sensors, gateways, vehicles and robots that acts on data with no network round trip.

  5. Architecture · Compute and edge

    Modular edge data centres and AI towers

    In Operations centre, the modular edge data centres and the AI tower in a box layer holds AI tower in a box, prefabricated and containerised modules and telecom and IT power systems, each drawn against the standard it follows. A site package that combines RAN, accelerated compute, storage, power, cooling and backhaul at a tower or remote site.

  6. Architecture · Compute and edge

    Kubernetes and fleet orchestration

    In Operations centre, the Kubernetes and fleet orchestration layer holds Kubernetes at the edge, fleet and multi-cluster management and bare-metal and zero-touch provisioning, each drawn against the standard it follows. Lightweight, conformant Kubernetes on single nodes and small clusters at remote sites.

  7. Architecture · Security, identity and trust

    Device identity and attestation

    In Operations centre, the device identity and attestation layer holds Trusted Platform Module, DICE layered identity and secure device identifiers, each drawn against the standard it follows. A hardware root of trust that holds keys, measures boot and signs attestation evidence.

  8. Architecture · Compute and edge

    Cooling and ruggedisation

    In Operations centre, the cooling and ruggedisation layer holds air cooling and containment, direct-to-chip liquid cooling and single-phase immersion cooling, each drawn against the standard it follows. Airflow design, containment and filtration for cabinets and small rooms. Cold plates and coolant loops that handle accelerator heat densities air cannot.

  9. Architecture · Compute and edge

    OTA updates, GitOps and software supply chain

    In Operations centre, the OTA updates, GitOps and software supply chain layer holds GitOps, secure update frameworks and device management, each drawn against the standard it follows. Declared desired state in version control that sites pull and reconcile, giving audit and rollback.

  10. হালনাগাদ

    Updates in rings, with rollback

    Software for Operations centre is released in rings, so a fault shows on one canary site long before it could reach the whole fleet. A ring for Operations centre goes ahead only on health and data completeness, and a failed gate rolls each node back without anyone travelling to site.

  11. Architecture · Compute and edge

    Storage and edge data management

    In Operations centre, the storage and edge data management layer holds NVMe and NVMe over Fabrics, object and file storage at the edge and storage security and media sanitisation, each drawn against the standard it follows. Low-latency flash storage locally and across the edge network.

  12. What you get

    Open where it can be, contracted where it must be

    What is open for Operations centre is free: the protocol, the methods and the open data can be used and built on by anyone. What must be contracted for Operations centre, sites, fleets, managed operations and data tiers, Nexus Observatory Inc. delivers at a price per site and per fleet, with service levels in numbers.

  13. Next step

    Operations centre, planned with you

    Tell us which layers of Operations centre you already run and which you want us to design and deliver. An engineer replies with a survey proposal, a design and a scope for Operations centre, and the plan grows from one site to the network you need.

    Plan a deploymentOpen the Field Deployment StudioSee Operations centre

More presentations

See it from every side.

Each opens as a full-screen presentation: deeper pages on the same subject, explainers, stories, guided tours of the places, network scenes and the observatories that came before.

All presentations

In depth

Use case Deployable incident observatoryUse case Avalanche detection siteISIC section · ISIC O Public administration and defence; compulsory social securityISIC section · ISIC N Administrative and support service activities

Explainers

Explainer · Government and public safety Government and public safetyWildfire cameras on real terrain, GBON station density, street-flood alert thresholds, CAP alerting with the authority in the loop and the radio horizon of a deployable cell.Explainer · Defence-adjacent (civil dual-use) Defence-adjacent (civil dual-use) — SectorsLawful civil dual-use sensing only: northern observatories, satellite reach at high latitude, maritime safety and ice, and critical-infrastructure resilience.প্ল্যাটফর্ম Operations centreAlarms nobody ownsপ্ল্যাটফর্ম Hardware & powerThe lab is kind; the site is not

Stories

Story · Resilient operations When the links failObserving systems matter most on the days infrastructure fails. Designing for that day means local autonomy, stored power, store-and-forward delivery and…Story · Airspace and drones A shared sky below 120 metresFlying beyond visual line of sight turns a drone from a camera on a stick into infrastructure. It needs a command link that holds, an identity others can…Story · Mountains and avalanche terrain Snowpack, slope and signalAvalanche forecasting reasons about the likelihood and size of avalanches on given terrain. The observations behind it come from a few stations in hard…

Places and networks

Guided tour · Guided 3D tour Wildfire forestCamera masts on the ridges, a weather station in the stand and a private cell: the red ground is what the ridge camera can see.Guided tour · Guided 3D tour Mountain valley siteA private 5G tower, an automatic weather station, a compact radar and a seismic vault on real alpine terrain, joined to an edge data centre that admits, times and publishes every…Guided tour · Guided 3D tour Arctic stationA self-powered station on the ice: a microgrid, a low-orbit satellite terminal, a weather station, a GNSS monument and an edge node that holds data through the polar night.Network scene · NTN · Satellite Satellite direct to device in a mountain valleyWhere the valley cell cannot reach, phones on the pass road connect straight to a low-orbit satellite, which relays through its feeder link to a gateway.Network scene · Backhaul · Fresnel zone Microwave backhaul across a river valleyAn 18 GHz hop from the church tower to a mast on the far side of the flood plain: its first Fresnel zone, clearance over the town and trees, and what a new building in the path…

Heritage observatories

Heritage observatory · United States · 1961 World-Wide Standardized Seismograph NetworkIn the early 1960s the United States put identical seismographs, calibrated the same way, in more than a hundred places around the world and published the records on film for…Heritage observatory · China · 132 Zhang Heng’s seismoscopeIn 132 CE a court astronomer presented a bronze urn that, it was recorded, told the capital of an earthquake no one in Luoyang had felt. No original survives.

Interactive tools

Interactive tool · Plan a site Field Deployment StudioTerrain, devices, coverage, viewshed, power and the best spot for a station, then a deployment request with the plan attached.

01Reference architecture

From one instrument to the grid.

Every deployment is drawn on the same six layers. A typed contract sits at every boundary, so parts from different makers fit together.

  1. 01InstrumentThe sensor or radio that makes the measurement, with its calibration.
  2. 02NodeThe device that hosts instruments: identity, time, power and storage.
  3. 03SiteNodes in one place, on one mast or in one shelter, with their links.
  4. 04CellSites that share a backhaul, a power plan and an operator.
  5. 05ClusterCells run as one service, with one fleet plan and one evidence trail.
  6. 06GridClusters that federate across owners through the open protocol.
01যোগ দিনIdentity, attestation quote, signed policy.02AdmitAdmitted for a stated scope, nothing wider.03বর্ণনা করুনWhat, where, how and by whom, with calibration.04TimeOne clock across the grid, with its error bound.05ObserveEvery value with quality and provenance.06QualifyTested against the specification, state on record.07FederateSigned exchange across grids and trust domains.08OrchestrateSoftware and configuration by signed plan.09CorrectA correction marks what depends on it and publishes a new version.10ExchangeProducts reach people and applications through versioned interfaces.11ProveSigned proof receipts that an independent verifier can check.12Exit and decommissionAn open exit path, and a record that closes.

The protocol functions this platform relies on most.

02Trust and evidence

Every observation carries its proof.

What a buyer, a regulator or a partner grid can check, on every site we build.

প্রমাণ

Identity and attestation

Hardware-rooted identity, with a quote checked at every join.

প্রমাণ

Segmentation

Least privilege between nodes, tenants and sites.

প্রমাণ

Time integrity

Every clock monitored, every holdover logged.

প্রমাণ

Evidence bundles

What was done, by whom, with what result.

প্রমাণ

Qualification state

Each device and release, with the tests behind it.

প্রমাণ

Signed plans

Software and configuration change only by a signed plan.

03Read further

The open specification behind it.

Observatory Protocol

Specification

Functions, state machines, objects and flows. Read →
Observatory Protocol

Reference designs

Nodes and sites drawn to the protocol. Read →
Observatory Protocol

সামঞ্জস্য

Run the suite against an implementation. Read →

04এই সাইটে

Where to go next.

Operations centre

ক্ষেত্রসমূহ

Every area, with its systems and standards. Open →
Operations centre

Technologies

The layered map of what we build with. Open →
Operations centre

Industries

The industry sites built on this platform. Open →
Operations centre

সম্পদ

Explainer, stories, specification and forms. Open →

Bring us the site.

We bring the survey, the design and the crew. The form names who receives your details.

Plan a deploymentযোগাযোগ

How operations centre is built.

Bring us the site.

We bring the survey, the design and the crew. The form names who receives your details.