Haystack / Project Haystack

Project Haystack is what makes building data usable at scale. Without consistent tagging, a thousand points across fifty sites is just noise; with it, analytics and dashboards can find “all the discharge-air temperature sensors” across the whole portfolio. Software Pile applies Haystack tagging in Niagara so your data is ready for what comes after controls.

Haystack Work We Do

  • Haystack tagging strategy and a tag dictionary that fits your equipment and analytics goals
  • Bulk tagging across stations using purpose-built Workbench tools – not thousands of manual edits
  • Tag validation and audits so the model stays consistent as sites are added
  • Preparing Niagara data for analytics platforms and digital-twin projects that expect a tagged model

Tagging Is an Investment in Everything After

Analytics, fault detection, energy optimization, and portfolio reporting all depend on a consistent data model. Tagging done right once pays off across every project that touches the data later. We do it at bulk scale and keep it consistent as the portfolio grows.

Tell us your site and point counts and what you plan to do with the tagged data. We will scope a tagging approach that serves it.

What Is Project Haystack?

Project Haystack is an open-source initiative that defines a standard way to describe building equipment and data using tags. Instead of relying on a point being called AHU3_SAT and everyone remembering that means supply air temperature on air handler 3, the point carries tags that state what it is, what equipment it belongs to, and what site that equipment is in.

The purpose is machine-readability. A tagged estate can be queried, analysed and visualised by software that was never told your naming convention — which is what makes fault detection, analytics and portfolio-level reporting practical across buildings that were engineered by different people in different years.

What Is a Haystack Tag?

A tag is a name, sometimes with a value, attached to an entity. Tags combine to convey meaning: a point might carry tags marking it as a sensor, as measuring temperature, as relating to discharge air, and as belonging to a particular air-handling unit, which in turn belongs to a site. No single tag says everything; the combination does.

The model also captures relationships — which equipment serves which space, which meter measures which system — so software can navigate the estate rather than just list points.

What Is nHaystack?

nHaystack is the Niagara implementation that exposes a station’s data over the Haystack API, letting external analytics and applications consume tagged Niagara data in a standard form. In practice it is how a Niagara estate gets connected to a third-party analytics platform without a bespoke integration per site.

N4 also has its own native tagging and relation model, which can be used alongside or mapped to Haystack tags. Which route fits depends on what is consuming the data at the other end — a question worth settling before tagging begins rather than after.

How Is Haystack Different from Brick Schema?

Both aim to describe building systems in a machine-readable way, and both have real adoption. Haystack grew from the controls and analytics industry and is tag-based; Brick comes from academic and research work and is expressed as a formal ontology. There has been convergence work between them.

For most Niagara estates the practical question is not which is theoretically better but which one the software you intend to buy actually consumes. Choose to match your analytics platform, not to win an architecture argument.

Do You Actually Need Haystack Tagging?

Not always. Tagging is an investment that pays off when something consumes it. It is clearly worth it if you are deploying analytics or fault detection, managing a portfolio where buildings must be compared, planning to integrate a third-party platform, or standardising engineering across a large estate.

It is harder to justify for a single building with a stable operator team, no analytics plans, and graphics that already work. Tagging an estate and then never connecting anything that reads the tags is a common and expensive way to produce no benefit.

Tagging an Existing Estate

Retrofitting tags across buildings engineered over many years is mostly a consistency problem rather than a technical one. Doing it by hand across thousands of points is slow and produces uneven results, which defeats the purpose. This is one of the strongest cases for custom Workbench tooling — bulk tagging driven by rules, with an audit report showing what was applied and what needs a human decision.

Related services: Niagara engineering tools and bulk tagging · BAS integration services.