Multi-sensor Indoor Positioning
Radio frequency, inertial motion and the environment's geomagnetic features contribute to an indoor trajectory on a smartphone. When the project calls for it, the architecture can extend to other sources, including the device camera.
Indoors, signals change with the space
NIST research on RSS describes indoor radio propagation as a difficult problem: walls, furniture and reflective surfaces create attenuation and multipath, while a radio map can become sensitive to changes in the site. RSSI remains a useful observation, but its strength does not map deterministically to distance.
An experimental study of body shadowing shows that the human body can attenuate Bluetooth signals and introduce errors in RSS-based systems. Device orientation, nearby people and moving obstacles therefore change the radio picture along a route.
The GiPStech article on Bluetooth RSSI translates these effects into an applied example and explains why beacons can provide useful references within a wider architecture. It is a company source and is not used as independent performance validation.
Three sources, three distinct roles
Each source observes a different aspect of motion or the environment. A camera-free configuration assigns each source a role consistent with what it can measure and with its limits, rather than treating the sensors as interchangeable.
Radio frequency
Beacons, networks or other signals can provide references tied to the space. Coverage, obstruction, multipath and orientation alter their quality.
Inertial engine
Accelerometers and gyroscopes describe steps, motion, direction and orientation. The estimate maintains local continuity but accumulates drift without corrections.
Geomagnetic fingerprint
Local magnetic variations within a building can become recognizable features when collected in a reference map suited to the site.
Sources contribute according to their available quality and their role in the configuration.
One shared estimate, not three parallel signals
A survey of pedestrian inertial systems finds that dead reckoning can support short- to medium-term tracking, but needs regular absolute fixes to constrain accumulated errors. Radio and geomagnetism can provide corrective observations with different properties.
Within the GiPStech scope, the camera-free engine interprets available inputs within one trajectory: inertial sensing follows motion between corrections, radio connects the estimate to references in the environment, and the geomagnetic fingerprint matches the route against mapped local features.
The Sensor Fusion for Localization page describes the wider architecture, which can also include vision and GNSS. The combination described here focuses on camera-free positioning.
Configuration changes what the system can observe
The three configurations represent different levels of information and project effort. They do not guarantee the same accuracy: coverage, continuity, infrastructure, sensor quality and site characteristics must be assessed together.
RF-only
Uses existing or dedicated radio references for localization whose granularity depends on technology, geometry, coverage and environment. On its own, it does not provide inertial trajectory continuity.
RF plus inertial
Adds steps, motion and orientation between radio references. It can make the trajectory more continuous, but drift and fix quality remain variables to control.
RF, inertial and geomagnetism
Adds an environmental reference based on a magnetic map of the site. It requires collection, validation and maintenance suited to materials, routes, devices and building changes.
A continuous trajectory needs credible corrections
Research on geomagnetic fingerprints highlights two central requirements: spatial differentiation and stability over time. Different buildings can produce very different fingerprints, while similar values may occur far apart; collection and validation must therefore represent the real site.
When an RF reference weakens, inertial motion can carry the trajectory for an interval and geomagnetism can provide a local check where the map is informative. Later corrections help constrain drift, but actual behavior depends on configuration, device posture, crowd density, coverage and mapping quality.
Mapping can use radio and magnetic data collected along representative routes with ordinary devices; duration, sample density and update cycles depend on context. Before defining the plan, the project must account for area accessibility, environmental variability and integration requirements.
Configuration starts with the environment and the required outcome
The GiPStech Technology page places radio, inertial sensing, geomagnetism and vision within the wider portfolio. The multi-sensor configuration described here can be extended by integrating the smartphone camera when the project calls for it. The choice still depends on the target, required granularity and operating conditions.
GiPStech Solutions show applications in airports, industry, healthcare and public spaces. These environments may benefit from camera-free positioning when privacy, device posture, lighting or continuity of use make vision unsuitable; every scenario still requires dedicated technical validation.
Before selecting a configuration, define the people or devices to locate, areas and floors, available infrastructure, expected behavior during signal degradation, mapping, integration and maintenance. Accuracy and continuity must be validated in the real context and do not transfer automatically between configurations.
Explore the wider architecture through Sensor Fusion, then compare it with Visual Positioning and VIO and adaptive hybrid localization.
Tell us about the environment and localization requirements
Share the targets, areas, available infrastructure, mapping needs and expected behavior. We can help define a camera-free configuration that can be verified in the field.
