Network Hero Animation — Light
Learn

Hub-and-Spoke vs. Point-to-Point Integration for Engineering Tools

Point-to-point integrations multiply with every tool you add. A hub-and-spoke digital thread connects each tool once and keeps them all aligned.

Every engineering program runs on a toolchain: requirements in one tool, system models in another, parts and BOMs in PLM, tests and simulations somewhere else. The data in those tools depends on each other. A requirement drives an architecture element, which drives a part, which drives a test.

And the toolchain is never fixed. Teams need the freedom to use the best tool for each job, the tools already in their organization’s ecosystem, and the tools a program or customer requires. Whatever connects them has to keep up as that mix changes.

The question is how to connect them. Most teams start with point-to-point integrations, one link per pair of tools. It works for two or three tools. Past that, the number of links grows faster than any team can maintain.

Hub-and-spoke takes a different approach: each tool connects once, to a central hub, and the hub keeps every connected tool aligned.

Point-to-point: one link per pair of tools

In a point-to-point setup, every pair of tools that needs to share data gets its own custom integration. Requirements to models is one link. Models to PLM is another. Requirements to test is a third.

The math works against you. Connecting every tool to every other takes n(n−1)/2 links:

ToolsPoint-to-point linksHub-and-spoke connections
333
5105
104510
1510515

The link count is only part of the cost:

  • Every link is custom. Each one maps one tool’s data structure to another’s, and that mapping is written and owned separately.
  • Upgrades break things. When one tool upgrades, every link touching it needs to be checked and often rebuilt.
  • Meaning drifts. Each link translates data its own way, so the same requirement can mean slightly different things in different places.
  • Impact stays hidden. A change only travels along the links someone built. Nobody sees the full chain from a requirement to the test that verifies it.

Hub-and-spoke: connect once, work with everything

In a hub-and-spoke model, each tool connects to a central hub through a single adapter. The adapter publishes the tool’s data to the hub and subscribes to the changes it needs from the hub. Tools never talk to each other directly.

  • One connection per tool. Ten tools means ten adapters, not 45 links.
  • Adding a tool is one step. A new tool gets one adapter and immediately works with everything already connected.
  • Upgrades stay contained. When a tool upgrades, only its adapter needs attention.
  • Teams choose their tools. Use the best tool for the job, the ones already in your ecosystem, or the ones your program requires. Engineers keep working where they know, and trace links they create there become part of the thread.

The hub becomes the place where relationships across tools live, so a change in one tool can be traced to everything it affects.

Side by side

Point-to-pointHub-and-spoke
Connections for 10 tools45 custom links10 adapters
Adding an 11th toolUp to 10 new links1 adapter
When one tool upgradesEvery link touching it is at riskOnly that tool’s adapter
What data meansTranslated differently in each linkMapped once to a shared model in the hub
Seeing a change’s impactOnly along links someone builtAcross every connected tool
Where cross-tool relationships liveScattered across integrationsIn one place: the hub

Why the hub needs an ontology, not just a data pipe

A hub that only moves data from one tool to another is still point-to-point underneath. The mappings just move inside the hub. Interoperability needs more: the hub has to understand what the data means.

That is the role of an ontology. It defines engineering concepts (a requirement, a system element, a part, a verification) and how they relate. Each adapter maps its tool’s data to those concepts once. After that, a requirement and the model element that satisfies it are understood as the same kinds of things, linked the same way, no matter which tools they came from.

  • Data keeps its meaning across tools and lifecycle stages.
  • Versions and configurations stay aligned, so you can see which revision of a requirement drove which revision of a part.
  • Change impact is traceable end to end, from a requirement through the architecture to the BOM and tests.
  • AI has trustworthy ground truth. An AI model reasoning over linked, versioned, well-defined data gives answers you can trace back to their sources.

SBE Vision starts from its Model Based Engineering (MBE) ontology, which is built on the Basic Formal Ontology (BFO), an ISO-standard top-level ontology (ISO/IEC 21838-2). MBE inherits BFO’s foundation and extends it with engineering concepts refined over years of real customer programs. The platform can also work with multiple ontologies, so the model fits the program rather than the other way around. Why this matters for AI →

How SBE Vision does it

  • The hub: the Digital Thread Platform holds the ontology, configuration management, and the applications that sit on the thread.
  • The spokes: Digital Thread Adapters, one per tool, each publishing to and subscribing from the hub. SBE Vision maintains and upgrades them.
  • Custom tools: the Adapter SDK connects legacy, in-house, and classified systems the same way.
  • AI on the thread: ThreadIQ™ works over the linked data, with autonomy matched to engineering consequence.

The platform runs in secure and classified environments, including air-gapped deployments for defense programs. See the Architecture page for the technical detail.

Frequently asked questions

Does hub-and-spoke replace our existing engineering tools?

No. Your teams keep working in the tools they use today. The hub connects them; it doesn’t replace them.

Can we change tools later?

Yes. Because each tool connects to the hub on its own, swapping a tool means connecting its replacement. The rest of the thread, and every other connected tool, stays in place.

Is a hub-and-spoke digital thread the same as an enterprise service bus?

No. A service bus moves messages between systems, but the meaning of the data is still translated pair by pair. A digital thread hub maps every tool’s data to a shared ontology, so relationships and meaning hold across all connected tools.

What happens when one of our tools is upgraded?

Only that tool’s adapter is affected. With SBE Vision, adapter maintenance and upgrades are handled by SBE Vision.

What if a tool we use doesn’t have an adapter yet?

The Adapter SDK connects it the same way as every pre-built adapter, including legacy, in-house, and classified systems.

Do engineers have to change how they create trace links?

No. Engineers create trace links in their native tools, and the platform brings those links into the digital thread.

Can hub-and-spoke work in classified or air-gapped environments?

Yes. SBE Vision is deployed in secure and classified environments, including air-gapped ones.

Connect each tool once. Work with everything.

Tell us which tools your program runs on, and we’ll show you what your digital thread looks like on a hub.