Project CassandraHET-B1
CassandraAll models
Models

Under development

HET-B1

Understand heat. With context.

A proposed heating-system intelligence programme for industrial heaters, boilers and heat pumps. Explore three cumulative capability levels and their planned architecture.

Explore the proposed levels

Three levels. One reviewed path.

Proposed capability levels. None is a released HET-B1 product.

Three cumulative proposed architecture layersObserve establishes measurement quality and equipment context. Investigate adds comparison and evidence. Anticipate adds forecasts, verification and approval. Human review applies at every level. MeasurementsQualityContextObserve ReferenceComparisonEvidenceInvestigate ForecastVerificationApprovalAnticipate
Proposed architecture. Human review at every level.

Proposed level

Build a trustworthy operating picture.

Organise available measurements around the equipment, energy input and useful heat output. Keep data gaps and uncertain values visible.

Planned architecture

  1. Measurements
  2. Quality checks
  3. Equipment context
  4. Operating view
  5. Human review

A clear operating view comes before a diagnostic conclusion.

Proposed level

Explain changing heating performance.

Compare similar operating conditions with an agreed reference. Present unusual behaviour, supporting measurements and alternative explanations for engineering review.

Planned architecture

  1. Observe foundation
  2. Performance reference
  3. Departure screening
  4. Evidence summary
  5. Human review

An unusual pattern would guide investigation, not establish a confirmed fault.

Proposed level

Plan with forecasts that can be checked.

Estimate future operating conditions and compare forecasts with later observations. Propose reviewed improvements while retaining previous versions.

Planned architecture

  1. Reviewed history
  2. Forecast candidate
  3. Chronological evaluation
  4. Approval
  5. Versioned release

Forecast quality would be reported after comparison with measured outcomes, without guaranteed future scores or fault dates.

Proposed architecture. Release depends on equipment-specific data, evaluation and engineering approval. Planned as advisory analysis; equipment protection and control remain separate.

Equipment context

Different heaters. Appropriate measures.

Performance measures depend on the equipment: heating COP for heat pumps, useful heat versus fuel energy for boilers, and useful heat versus electrical input for electric heaters.

Equipment type, available measurements and the energy boundary would be established before analysis. A single efficiency measure cannot describe every heating technology.

Compare the programmes Try the synthetic heating demo