Technical Trade SchoolBAS-400 developer-series
BAS-403
Ten concept and evidence figures

BAS-403 Visual Aid Library

Figure-use rule

Each figure is an orientation and reasoning aid, not a project drawing. Learners must pair it with the current sequence, point list, wiring/piping documents, manufacturer instructions, and as-found evidence.

Figure 1 — Information-model purpose, users, and identity critical

Normal-rule anchor

A name is a label; identity and meaning require stable keys, context, units, relationships, quality, and governance.

Case overlay

Three front-end paths use different names for the same discharge-air sensor, while two identical labels refer to different physical sensors.

Calculation/evidence callout: Reconcile source records to unique physical and digital identities; duplicates, aliases, orphans, and unresolved collisions must be counted.

Figure 2 — Naming conventions, dictionaries, and revision control

Normal-rule anchor

A convention must define grammar and governance, not merely provide examples; abbreviations and exceptions belong in a controlled dictionary.

Case overlay

A controller limits object names to sixteen characters while the enterprise standard requires readable multi-building uniqueness.

Calculation/evidence callout: Calculate maximum lengths, collision counts, abbreviation coverage, and migration impact across supplied target constraints.

Figure 3 — Point metadata, units, ranges, facets, and quality critical

Normal-rule anchor

A numeric value without units, range, timestamp, quality, and source can be plausible and still be unsafe or meaningless.

Case overlay

A temperature point displays 72.0 with no unit, a 0-10 V raw range, and a stale quality flag hidden by the graphic.

Calculation/evidence callout: Verify engineering-unit conversion, range, precision, age, and plausibility; count missing or conflicting required facets.

Figure 4 — Equipment, location, and relationship hierarchy

Normal-rule anchor

Containment, service, flow, and control are different relationships; forcing them into one folder tree destroys meaning.

Case overlay

A VAV controller is physically in a ceiling, belongs to an air system, serves a zone, and is commanded by a central schedule.

Calculation/evidence callout: Test graph completeness with counts and queries for orphan points, equipment without spaces, zones without serving equipment, and broken upstream/downstream paths.

Figure 5 — Project Haystack tags and references

Normal-rule anchor

Tags describe meaning in combinations; adding many tags without a controlled convention can create contradictory or nonportable models.

Case overlay

Legacy VAV points contain names but no equipment references, and airflow values mix sensor, setpoint, and command meanings.

Calculation/evidence callout: Calculate tagging coverage, valid-reference rate, query precision/recall on the supplied answer set, and exception count.

Figure 6 — Brick entities, classes, relationships, and validation

Normal-rule anchor

Ontology modeling separates what an entity is from how it relates; a valid label does not compensate for a wrong class or missing relationship.

Case overlay

An AHU discharge sensor is modeled as a generic temperature point with no equipment or location relationship.

Calculation/evidence callout: Count validation violations, classify missing versus conflicting relationships, and test required competency queries against expected results.

Figure 7 — BACnet object identity and semantic crosswalk critical

Normal-rule anchor

BACnet object name is not a globally unique enterprise identity; network/device/object context and mapping provenance must be preserved.

Case overlay

Two vendor devices expose identical object names and one gateway changes engineering units during mapping.

Calculation/evidence callout: Construct unique keys, verify object counts, reconcile units/enumerations, and identify duplicate, missing, or unstable identifiers.

Figure 8 — Templates, normalization, import, and mapping critical

Normal-rule anchor

Bulk automation multiplies both good rules and bad assumptions; import requires validation before and after every controlled batch.

Case overlay

A CSV import would rename 4,000 points, assign units by suffix, and overwrite existing descriptions.

Calculation/evidence callout: Calculate match, exception, collision, overwrite, and rollback counts; verify idempotence on a second dry run.

Figure 9 — Data quality, time, scaling, status, and lineage critical

Normal-rule anchor

Data quality is use-specific; a value acceptable for an operator overview may be inadequate for control, fault detection, billing, or compliance.

Case overlay

A historian resamples five-minute COV data into one-minute values and an analytics rule treats interpolated values as direct measurements.

Calculation/evidence callout: Calculate age, completeness, clock offset, unit/scaling error, interpolation share, and quality-rule pass rate.

Figure 10 — Governance, validation queries, migration, and handoff critical

Normal-rule anchor

Semantic work is configuration management: every change can affect graphics, trends, alarms, analytics, integrations, and human search.

Case overlay

A naming cleanup improves consistency but breaks saved reports and external API consumers that rely on legacy paths.

Calculation/evidence callout: Reconcile affected consumers, validation results, migrated/deferred/failed entities, aliases, and open exceptions before release.