Skip to Content

Hong Kong · founded 2025 · a start-up

Intelligent sensor interaction
starts with measurement.

We are researching a measurement model within the sensor to analyse signals and support reading assessment alongside physical calculations. Our first PoE smart flow sensor is in development; a planned SaaS + AI cloud engine will extend this work to cross-sensor reports and risk forecasting.

Measurement model: design and simulation tests · first product: system design

Our view

AI belongs at the measurement signal layer.

A reading must be assessed against the conditions in which it was measured. Changes in the medium, temperature or installation may affect the signal and the result. Our research focuses on learned models for signal analysis and reading assessment under complex operating conditions.

  • Medium

    Bubbles or particles may affect signal quality.

  • Temperature

    Temperature changes may alter fluid properties and the conditions for measurement.

  • Installation

    Bends, valves and installation position may influence the flow at the measuring point.

Progress in learned models for language and images brings new tools to measurement research. We believe the greater opportunity for AI in measurement lies at the signal layer, before a reading is produced; our planned cloud analysis will build on the outputs of that layer.

Our approach

From signal analysis
to system intelligence.

Our roadmap links three layers: a measurement model within the sensor, PoE connectivity, and a cloud engine organised around system topology. The ambition is to extend AI from individual measurements to reports and risk forecasts for connected physical systems. Each layer is at a different stage.

  1. Signal layer

    Measurement model

    Research aim: a measurement model that works alongside physical calculations to analyse signal patterns and support reading assessment under complex conditions. The model is currently in design.

    Research
  2. Connection layer

    PoE, one cable

    In development: a PoE system and data platform, aiming to carry power and data over one network cable. Access to raw signals from the measuring instrument has not yet been confirmed with the manufacturer.

    Being built
  3. System layer

    Analysis across sensors

    Planned: a SaaS + AI cloud engine linking flow, infrared and other sensors through system topology, using AI agents for analysis, reports and risk forecasting. Not started.

    Planned

First product

PoE smart flow sensor

Our PoE smart flow sensor is the first product in development along this path.

  • Measuring body: a proven flow meter from a partner manufacturer.
  • Our development: the PoE system and data platform are being built; the measurement model within the sensor is in research.
  • First application direction: pipe flow monitoring.

In development

More on the flow sensor
Concept picture: a sensor fitted on the outside of a water pipe, connected by one network cable to a small switch on the wall.
Concept picture (generated), not a photograph of the product.
Flow
0.80 m³/h
Signal quality
Normal
Formula and model
Consistent

Example data, not the output of a real device.

Our foundations

Our platform is in use; the measurement model is in design and simulation testing.

  • Company

    Incorporated in Hong Kong

    SENSINGNET LIMITED is a private company incorporated in Hong Kong. Its registered name, company number, date of incorporation and registered office can be checked on the company page.

    Company details
  • Engineering

    An internal platform already in use

    Our own platform supports company projects, a customer portal and business records, with access-controlled data interfaces for AI. It provides practical experience in permissions and business integration for the sensor data platform now being developed.

  • Research

    Model design and simulation tests

    The measurement model is in design; a first round of tests on simulated data is complete. Collecting real signals is next and has not started. Validation on real data will follow.

Roadmap

Development milestones

Completed work is dated. Work ahead shows its sequence and the conditions for the next stage, without dates.

  • April 2025

    Company incorporated in Hong Kong

    Done
  • Autumn 2026

    Own management platform and AI interface in use

    Done

Research line

Measurement model

  1. In progress

    Model design and simulation tests

    Design the model framework and test it on simulated data; a first round of tests is complete. Move to the next stage once the framework is settled.

  2. Next

    Collection of real signals

    Record raw sensor signals on a test pipe with independent reference readings. Not started.

  3. Later

    Validation on real data

    Compare the model with conventional calculations against independent references; return to design if the results do not hold.

  4. Later

    Model embedded in the sensor

    Compress the validated model to run on the sensor processor.

Product line

Sensor and platform

  1. In progress

    System design of the first PoE smart flow sensor

    Define the sensor system, its PoE connection and data platform.

  2. Next

    Demonstration system

    Connect sample units to our platform to demonstrate the system.

  3. Later

    First pilots in Hong Kong

    After the demonstration system, agree the scope and acceptance method for each pilot site.

  4. Further out

    More sensor types on one network

    Extend the platform to other types of sensors.

The two lines advance side by side. The measurement model goes into the sensor only after it has been validated on real data.