Lighting is increasingly becoming a distributed source of building intelligence. Recent connected-lighting work makes the case for collecting better information at the point of use, but engineering value appears only when the information survives the path from device to decision to actual luminaire behavior. That path needs a defined installation context, a named control owner, and a record of what each component is known to do.
This article uses the current product record as a practical starting point. M3 provides Bluetooth 5.0, 24–24.25GHz sensing, up to 6m motion detection and up to 4m presence detection. Those parameters describe the device itself; they do not prove Bluetooth NLC, HVAC Integration NLC Profile, or gateway interoperability. The selection method is deliberately conservative: use published parameters to constrain a design, and use documented integration information plus site testing to validate every behavior that is not explicitly stated.
The engineering method is discussed against real models. Diagrammed examples and conceptual connections name their operating boundaries. Confirm product parameters against the current model, variant, and applicable technical documentation.
Why shared occupancy data matters
Lighting is increasingly becoming a distributed source of building intelligence. A project should begin with the operational question rather than a feature name. Identify the people, equipment, light level, response time, and maintenance outcome that the space needs. This makes it possible to distinguish a useful requirement from a marketing label.
A Bluetooth version is a radio capability, not proof of a particular application profile or building-management integration. The gateway, event model, firmware version, device health reporting, and receiving controller all require confirmation. Treat this distinction as a design input. It prevents a system diagram from assigning responsibilities to a product before its actual interface, documented function, and configuration route have been verified.
Keep the evidence attached to the decision that it supports. Record the model and variant, physical location, configuration revision, wiring or network condition, date, observer, measured result, and any exception. When a result changes after a setting, installation, or firmware change, preserve both records. This gives the next engineer a way to distinguish a changed requirement from a changed system response.
What the installed sensor actually proves
M3 provides Bluetooth 5.0, 24–24.25GHz sensing, up to 6m motion detection and up to 4m presence detection. Those parameters describe the device itself; they do not prove Bluetooth NLC, HVAC Integration NLC Profile, or gateway interoperability. Read each field together with its condition and unit. A current setting, voltage window, maximum mounting height, standby figure, or IP rating has meaning only in the context of the specific product variant and installed assembly.
Create a short “known / confirm / test” table for the project. Published data belongs in known; exact firmware, wiring, protocol, mounting construction, load, and operating condition belong in confirm or test until evidence moves them into the known column.
Keep the evidence attached to the decision that it supports. Record the model and variant, physical location, configuration revision, wiring or network condition, date, observer, measured result, and any exception. When a result changes after a setting, installation, or firmware change, preserve both records. This gives the next engineer a way to distinguish a changed requirement from a changed system response.
Define occupancy as useful building data
Start with a single representative room. Map each sensor to a room identifier, observe raw activity and presence events, then compare them with a manual occupancy log before allowing those events to change HVAC operation. Keep the first test deliberately small, with documented versions and a stable environment. A controlled pilot produces evidence that can be understood later; a full installation with several simultaneous changes often produces symptoms without a clear cause.
For every input, write the expected result and the owner of the decision. A sensor can report an observation, a controller can apply policy, and a driver or other actuator can execute a result. When two devices can issue conflicting decisions, define priority before the system reaches site.
Keep the evidence attached to the decision that it supports. Record the model and variant, physical location, configuration revision, wiring or network condition, date, observer, measured result, and any exception. When a result changes after a setting, installation, or firmware change, preserve both records. This gives the next engineer a way to distinguish a changed requirement from a changed system response.
Build an observable integration chain
Test occupied, temporarily vacant, quiet seated, disconnected, gateway-restarted, and recovered states. Lighting may use a fast response, while HVAC usually needs a different confirmation period and equipment-safe fallback. Run the scenarios in a fixed order and keep raw observations where possible: measured light, input power, device status, event time, configuration revision, and room condition. One convincing demonstration is not an acceptance method; repeatability is.
Fault behavior deserves the same attention as normal behavior. The system should state whether information is current, stale, missing, or invalid, then apply the approved fallback. A communications failure must not silently masquerade as a normal vacant or healthy state.
Keep the evidence attached to the decision that it supports. Record the model and variant, physical location, configuration revision, wiring or network condition, date, observer, measured result, and any exception. When a result changes after a setting, installation, or firmware change, preserve both records. This gives the next engineer a way to distinguish a changed requirement from a changed system response.
Pilot the controls before scaling
Use results to refine the zone, operating policy, and maintenance plan rather than to force a single universal setting. Installation changes, furnished spaces, varying daylight, user behavior, and electrical conditions often matter more than a nominal figure quoted without its test boundary.
Where an external system is involved, preserve an explicit interface contract: device and firmware identification, signal or data meaning, addressing, timeout rule, safety boundary, and recovery behavior. That contract is what keeps a working pilot from becoming an opaque deployment.
Keep the evidence attached to the decision that it supports. Record the model and variant, physical location, configuration revision, wiring or network condition, date, observer, measured result, and any exception. When a result changes after a setting, installation, or firmware change, preserve both records. This gives the next engineer a way to distinguish a changed requirement from a changed system response.
Hand over a maintainable data model
The finished system should make each state explainable: where it originated, how long it remains valid, who interprets it, and what happens when data is late or unavailable. Handover should include drawings, configuration exports, parameter selections, tested scenarios, the original site conditions, and a short sequence a technician can repeat after replacement or refurbishment.
The purpose of the document is not to claim that technology eliminates engineering judgement. It is to make that judgement visible, testable, and easier to carry forward as the project, the building, and the installed product set evolve.
Keep the evidence attached to the decision that it supports. Record the model and variant, physical location, configuration revision, wiring or network condition, date, observer, measured result, and any exception. When a result changes after a setting, installation, or firmware change, preserve both records. This gives the next engineer a way to distinguish a changed requirement from a changed system response.
Products and key parameters
These fields come directly from the current product library. Open the product page to confirm complete parameters and the applicable product record.
Standalone SensorsM3View full product parameters - Power supply
- Lithium manganese CR2477 button battery (original battery can be purchased at the Mindray JD.com/Tmall flagship store)
- Battery life
- 3 years
- Installation height
- Wall installation 1.5-2m
- Wireless protocol
- Bluetooth 5.0
- Radar frequency
- 24G Hz-24.25G Hz ISM band
- Detection distance
- Motion sensing 6m (maximum), presence sensing 4m (maximum)
- Detection angle
- Horizontal FOV about 130°
References and editorial note
Sources were reviewed on 2026-09-07. Recent industry sources and basic technical references are listed separately. The article is original engineering commentary grounded in the current product record, not a translation of the cited source. Any calculation or recommendation does not represent completed product interoperability testing.
