Asset Intelligence is the Homes Platform integration for asset-centric environmental stewardship, risk evaluation, provenance, and explainable home intelligence.
It extends Home Assistant with a new concept:
👉 Assets — the things your home exists to support, protect, and preserve
This transforms your home from:
monitoring devices
into
understanding what matters
To understand how the system works:
- Wiki: https://github.com/tom-tagmdl/asset_intelligence/wiki
- Getting Started: https://github.com/tom-tagmdl/asset_intelligence/wiki/Getting-Started
- Troubleshooting: https://github.com/tom-tagmdl/asset_intelligence/wiki/Troubleshooting
- Open HACS in Home Assistant.
- Go to Integrations.
- Click the menu and select Custom repositories.
- Add
https://github.com/tom-tagmdl/asset_intelligencewith categoryIntegration. - Search for
Asset Intelligenceand install. - Restart Home Assistant.
- Go to Settings > Devices & Services > Add Integration and add
Asset Intelligence.
In the integration options, configure:
- Default labels: for example
Asset - Document storage path: for example
/media/asset_intelligence_documents - Enable document management:
on
Notes:
- Default labels are applied automatically to new assets.
- The panel now prompts for a reload when a newer frontend bundle is detected after an upgrade.
- If frontend updates still are not immediately visible after an upgrade, perform a hard browser refresh.
Track this repository directly in HACS:
Asset Intelligence enables your home to:
- Understand what it contains (assets as first-class entities)
- Evaluate whether conditions are appropriate
- Track custody, movement, and loan history
- Maintain documents, provenance, and insurance records
- Provide explainable risk and advisory insights
- Support room-aware and audience-aware experiences
Today’s smart homes answer:
- “What is the temperature?”
- “Is the light on?”
- “Which device is in this room?”
But not:
- “Are conditions appropriate for what’s here?”
- “Is anything at risk?”
- “Where is this, and who has it?”
- “Do I have documentation for this?”
Asset Intelligence enables your home to answer those questions.
Asset Intelligence introduces a simple, powerful model:
- Areas (Rooms) → where things are
- Assets (Devices) → what exists
- Labels → what kind of things
- Environment → what’s happening
- Risk & Advisory → what it means
Sensors → Environment → Evaluation → Risk → Advisory → Activity → Automation
- Sensors capture conditions
- The system evaluates those conditions against asset requirements
- Risk is determined
- Advisory explains what matters
- Activity records what changed
- Automations (optional) take action
An asset is anything that matters.
- TVs, Sonos, networking gear, appliances
- Artwork, instruments, furniture, collections, documents
Each asset is implemented as a Home Assistant device, enriched with:
- environmental requirements
- documentation and provenance
- custody and governance
- history and audit data
Asset Intelligence brings concepts from:
- museums
- archives
- collections
- estate inventories
into the home.
It introduces:
- structured asset inventory
- environmental stewardship
- audit and lifecycle tracking
- custody and loan workflows
- documentation and provenance
This is not just automation.
👉 It is stewardship
Asset Intelligence models the environment in a structured way across:
- climate
- light
- air quality
- particulates
- biological conditions
- structural context
This allows the home to evaluate:
- Is this environment safe for artwork, instruments, electronics?
- Is air quality healthy?
- Is CO₂ elevated?
- Is the room becoming unsafe?
Room Detail now includes a room-scoped risk/advisory interpretation layer designed for explainability.
What this adds:
- State and Confidence context for the room snapshot
- Status since time tracking for the current condition state
- Clear reasons for the current room condition
- Advisory-oriented guidance tied to observed room signals
Why this matters:
- Users can understand room-level safety context before reading raw sensor values.
- It provides a faster decision path for placement and remediation actions.
Room Detail now includes a dedicated Room Confidence Drivers panel that explains exactly why the room confidence is what it is.
What it shows:
- Coverage: how many configured signals are actively reporting data (e.g. 4/5 — 80%)
- Configured signals: total count of sensors mapped to this room
- Signals reporting: count of sensors returning live values
- Signals missing: count of configured sensors with no data
- Confidence summary: a plain-language explanation of why confidence is Good, Partial, Degraded, or Stale
- Missing signal names: which specific sensor slots are empty
- Reporting signal names: which slots are actively contributing
Quick fix link:
- A Configure sensors link next to the missing signals list navigates directly to Room Configuration to fix gaps without manual navigation.
Why this matters:
- Users often see "Partial" or "Degraded" confidence with no clear cause.
- This panel surfaces exactly which sensors are missing data so users can assign or fix them.
- Confidence is not just a status — it reflects how much real-world evidence backs the room's risk and people-health assessments.
Room Detail now includes a dedicated People Health panel that evaluates room conditions against a baseline human profile.
MVP inputs:
- temperature
- humidity
- CO₂
- PM2.5
- VOC
MVP output:
- Health state (Green / Amber / Red)
- Confidence (High / Medium / Low)
- Status since
- top reasons and advisory guidance
- observed vs total signals coverage
Profile model (MVP):
- System default baseline profile in storage
- Optional room-level profile override support
- Preferred range and hard-limit checks per metric
Asset Intelligence includes a focused measurement workflow on the Asset Detail page.
- Start measurement begins a time-bound observation session for a specific asset in its current room context.
- While active, the header shows a live measurement box with:
- elapsed timer
- coordinator update count (how many new environment refresh cycles were captured)
- quick Stop measurement action
- Stop measurement requests session completion and finalizes the measurement profile from captured observations.
Use Start/Stop Measurement when you want to understand how a room behaves over a real observation window for one asset, not just a single instant reading.
Examples:
- Verify how stable temperature/humidity are around an instrument during evening hours.
- Capture changing light exposure for an artwork near windows.
- Compare before/after behavior while HVAC, shades, or room setup are adjusted.
Measurement sessions are written into the Activity pane with dedicated entries:
- Start entry: records session start timestamp and initial room-environment snapshot.
- Stop entry: records completion timestamp, observation/update count, and finalized measurement profile results.
This gives a clear audit trail of when the measurement began, how much data was collected, and what the final observed baseline looked like.
Room Detail now includes a dedicated Room Measurement History panel so users can evaluate room conditions over time and decide whether the room is appropriate for a given asset.
What this provides:
- A room-scoped timeline of measurement sessions across assets in that room
- Newest sessions at the top, oldest at the bottom
- Asset context on each entry (which asset the session was run for)
- Filter options for All, Start, Stop, and Asset
- A quick trend summary showing:
- session count
- last session timestamp
- average observations per completed session
Why this matters:
- You can compare room behavior across multiple sessions, times of day, and assets.
- You can validate whether a room remains stable enough for sensitive assets (for example, artwork, instruments, or electronics).
- You can make placement decisions using observed room behavior, not just one-point-in-time readings.
How this works with People Health:
- The People Health panel gives immediate room-health context.
- Room Measurement History provides the time-based evidence behind that context.
- Together they support both snapshot decisions and trend-based validation.
- Assets define what conditions matter
- Rooms define how conditions are measured
- The system compares the two
“Are these conditions appropriate for what is in this room?”
Instead of alerts alone, the system provides:
- risk states (Green / Amber / Red)
- advisory explanations
- missing data warnings
- event history
This creates:
👉 awareness, not noise
Each asset can hold:
- receipts and appraisals
- manuals and warranties
- insurance references
- provenance records
- physical document locations
Documents are stored in external storage (NAS recommended) and linked into the system.
Track:
- ownership status
- loans and returns
- custody transfers
- storage location
- agreements and responsibility
Asset Intelligence uses a layered architecture:
- Environment → sensing
- Evaluation → decision engine
- Coordinator → runtime state, events, advisory
- Entities → read-only projections
👉 See: ../../wiki/Architecture-Overview
Asset Intelligence enables:
- room-aware voice interactions
- audience-specific descriptions
- guided experiences
- contextual explanations
The home can eventually explain:
- what is in a room
- why it matters
- what is happening to it
Asset Intelligence is built around:
homes that understand what matters and act accordingly
Not just:
- convenience
- automation
But:
- awareness
- responsibility
- explanation
Asset Intelligence transforms Home Assistant into:
- an environmental awareness system
- an asset management system
- a documentation system
- a stewardship system
The result:
A home that behaves with intention.
| Path | Purpose |
|---|---|
__init__.py |
Integration entry point. Registers all services, sets up config entries, and wires the coordinator, storage, and document view into HA. |
manifest.json |
HA integration metadata: domain, version, config flow flag, IoT class, and codeowner declaration. |
quality_scale.yaml |
Tracks which Home Assistant Quality Scale rules have been satisfied for Bronze → Platinum progression. |
const.py |
Shared constants used across the integration (domain name, signal names, service names). |
models.py |
Core data models: Asset, CustodyRecord, LoanRecord, and supporting types. |
coordinator.py |
AssetIntelligenceCoordinator — owns runtime state, triggers refreshes, computes room-level human-health projections, and dispatches change signals to entities. |
asset_entity.py |
Base entity class shared by sensors and binary sensors; resolves the asset from coordinator data. |
sensor.py |
Sensor entities: asset count, asset list, room environment readings, and room-level people-health attributes. |
binary_sensor.py |
Binary sensor entities: risk state and advisory flags per asset. |
config_flow.py |
UI-driven config flow for setting up and reconfiguring the integration. |
storage.py |
AssetStore — reads/writes persistent JSON storage, including system defaults and human-health profile baselines. |
document_models.py |
Data classes for document records (receipts, appraisals, manuals, provenance). |
document_storage.py |
DocumentStorage — manages file-level operations for linked asset documents on NAS/share. |
environment.py |
Parses and normalises room environment sensor data into structured readings. |
evaluation.py |
Evaluates environmental conditions against per-asset requirements and room-level human-health profiles to produce risk/advisory states. |
advisory.py |
Generates human-readable advisory text from evaluation results. |
validation.py |
Input validation helpers used by service handlers to reject malformed payloads early. |
panel.py |
Registers the custom frontend panel with HA's HTTP layer. |
strings.json |
Translatable UI strings for the config flow and services. |
services.yaml |
HA service schema declarations (fields, descriptions) for all registered services. |
translations/en.json |
English translations for config flow, services, and entity states. |
helpers/document_resolver.py |
Resolves document paths and validates file availability against configured storage. |
services/document_retrieval.py |
Service handler for document fetch/download operations. |
frontend/panel_v5.js |
Compiled frontend panel served to the HA UI for asset management views. |
frontend/src/ |
Source for the frontend panel (components, pages, views). |
docs/asset_intelligence.md |
End-user documentation: concepts, installation, removal, and supported actions. |
docs/asset_intelligence_examples.md |
Usage examples and automation patterns for common asset intelligence scenarios. |
docs/asset_intelligence_troubleshooting.md |
Troubleshooting guide for common setup and runtime issues. |
docs/quality_scale_audit.md |
Platinum audit log: evidence mapping for each Quality Scale requirement. |
custom_components/asset_intelligence/brand/ |
Integration branding assets (icons, banners, logos) for the HA brand registry. |
tests/conftest.py |
Shared pytest fixtures for all test modules. |
tests/test_config_flow.py |
Tests for UI config flow: setup, validation, and single-entry enforcement. |
tests/test_setup.py |
Tests for integration startup and platform loading. |
tests/test_unload.py |
Tests for config entry unloading and resource cleanup. |
tests/test_services.py |
Tests for service handlers: document upload, attach, delete, and asset mutations. |
tests/test_documents.py |
Unit tests for document record helpers and rebuild logic. |
tests/test_history.py |
Tests for asset history payload construction. |
tests/test_diagnostics.py |
Tests for the HA diagnostics payload output. |
tests/test_entity_metadata.py |
Tests for sensor and binary sensor entity metadata (name, unique ID, device class). |
tests/test_validation.py |
Tests for input validation helpers. |
tests/run_quality_gate.ps1 |
PowerShell script that runs the full pytest suite with coverage gate (≥85%) and emits artifacts. |
tests/run_smoke_tests.ps1 |
Lightweight smoke run targeting the most critical service tests only. |
tests/_validate_services.py |
Standalone script to validate services.yaml schema against registered service definitions. |
QUALITY_SCALE_CHECKLIST.md |
Working checklist tracking Bronze → Platinum progress with phase-by-phase implementation plan. |