Navigate through changing GPS.
Rahbar continues pedestrian motion processing when GPS quality declines, then uses valid location updates according to their quality to correct the fused estimate.
GPS / MOTION FUSION RAHBAR / PEDESTRIAN NAVIGATION SDK
Keep people
moving when
GPS struggles.
Add GNSS-resilient pedestrian positioning to your iOS app. Rahbar combines inertial sensing, learned motion models, and uncertainty-aware fusion to maintain position estimates through naturally degraded GPS conditions.
iOS SDK · Offline-first · Geodetic and relative positioning.
OUT THERE, EVERY STEP MATTERS.
THE EXPERIENCE IS YOURS.
The positioning
is Rahbar.
FOR TEAMS BUILDING Hiking apps Trail experiences Outdoor navigation Indoor & field operations
EVALUATE RAHBAR
Test Rahbar on
your own iPhone.
Request complimentary access to the Rahbar iOS evaluation app for 15 days. Compare Rahbar’s fused position with the iPhone’s native location estimate during real walks in GPS-degraded places. The app does not simulate outages; evaluate natural conditions such as tree cover, narrow corridors, and building obstruction.
iPhone app · iOS 17 minimum
Current development build: iOS 26.6.2
Current development build: iOS 26.6.2
01 / THE SDK
A pedestrian navigation core.
Your product on top.
Rahbar is an on-device positioning stack for iOS that combines Core Motion, GPS, deep-learning motion inference, and uncertainty-aware position fusion. Your app receives fused position estimates with uncertainty and quality information.
Keep uncertainty visible.
Position output includes horizontal, directional, and vertical uncertainty plus quality and degradation states, so your app can make informed decisions.
QUALITY-AWARE OUTPUTOn-device positioning stack.
Rahbar provides positioning without an account or network connection. Your app provides the maps, routes, and user interface. Rahbar does not upload coordinates, tracks, or diagnostics.
ON-DEVICE POSITIONINGFor SDK integrations: A host app can disable GPS fusion and provide a manual starting point when no GPS position is available. This option is for app integrations; it is not exposed in the Rahbar evaluation app.
02 / UNDER THE HOOD
Sensor input in.
Fused position out.
Rahbar combines iOS Core Motion data with GPS, deep-learning motion inference, and uncertainty-aware position fusion.
Auto Gait gathers the walking data used to fine-tune Rahbar’s Deep Motion Model, helping the SDK adapt its motion estimates during evaluation.
The host app may also supply independent absolute or relative position measurements. Rahbar applies them according to their stated uncertainty rather than treating them as hard teleports.
Request a Rahbar evaluationFROM SENSOR INPUTS TO APP EXPERIENCE
Core MotionGPSHost measurements
Rahbar SDKGPS + motion fusion and Deep Motion Model
COREYour iOS appFused position + uncertainty + motion
Rahbar owns acquisition, normalization, fusion, inference, uncertainty, calibration, and session tracks. Your app owns the user experience.
03 / INTEGRATION
One framework.
Two position modes.
Add the signed XCFramework to an iOS app, choose how position should be expressed, and consume fused updates through a single long-lived RahbarNavigation instance.
iOS 17+Swift APIOffline navigationNo account requiredPhysical iPhone for field testing
OUTDOOR / GEODETIC
Latitude and longitude
Use fused geographic coordinates for trail, urban, and map-based navigation. Start from the first valid GPS fix or a known coordinate.
INDOOR / RELATIVE ENU
North, east, and down
Start from a relative origin and render meter-based output against a floor plan. Independent relative measurements can refine the estimate.
Add and authorize
Add RahbarSDK.xcframework as Embed & Sign, import RahbarSDK, and provide Location and Motion usage descriptions.
Choose the configuration
Select .geodetic or .relativeENU, the initial GPS-fusion state, and whether active navigation may continue in the background. An integrating app can disable GPS fusion and start from a manually supplied point when GPS is unavailable; the Rahbar evaluation app does not expose this control.
Run the lifecycle
Create one RahbarNavigation, request authorization, call start(), then begin a session from automatic GPS, a manual coordinate, or a relative origin.
Consume fused updates
Use position for navigation, including its uncertainty and source. Use motion, calibration, lifecycle, and quality fields to explain current behavior. Keep rawGPS diagnostic-only.
SWIFT / OUTDOOR PILOTRahbarSDK
import RahbarSDK
let config = RahbarConfiguration(
positionOutputMode: .geodetic,
gpsFusionEnabled: true,
allowsBackgroundLocationUpdates: true,
showsBackgroundLocationIndicator: true
)
let navigation = try RahbarNavigation(
configuration: config
)
_ = navigation.requestAuthorization()
Task {
for await update in navigation.updates {
let fused = update.position
let gps = update.rawGPS // diagnostics only
let quality = update.quality
}
}
try navigation.start()Field-test on a physical iPhone.
The Simulator can validate API integration, but it cannot reproduce real motion, GPS, uncertainty growth, background behavior, or battery use.
PILOT EVALUATION TOOLS
See what fusion changes.
Compare estimates
View Rahbar's fusion position session on a map to evaluate its accuracy.
Fine-tune with Auto Gait
See how Rahbar's Auto Gait feature fine-tunes the Deep Motion Model used by Rahbar’s pedestrian navigation core.
Add known observations
Inject absolute or relative measurements with realistic uncertainty. Rahbar weights the correction instead of snapping to it.
04 / TAKE IT OUTSIDE
Start with a walk.
See where it leads.
RAHBAR APP IN ACTION
Request a 15-day TestFlight evaluation and compare Rahbar with native iPhone location during natural GPS degradation. Browse selected trail, urban, and store sessions below.
Request an evaluation
TRANSPARENT, SIDE-BY-SIDE EVALUATION
Compare Rahbar with native iPhone location.
This iPhone 15 Pro capture shows Rahbar’s blue position and uncertainty beside the native iPhone location estimate in red. It was recorded after GNSS initialization during a naturally degraded walk through a narrow corridor bordered by four-story buildings and trees. The app lets evaluators inspect both solutions together; no outage is simulated.
The native location estimate is a comparison signal, not independent ground truth. Map alignment alone does not establish measured accuracy.
PRELIMINARY DISTANCE-RELATED DRIFT ESTIMATE~4%of distance traveled
Based on approximately 60 hours (about 300 km) of data collected during GPS outage and degradation conditions. At this rate, a 1 km walk corresponds to about 40 m of accumulated position drift. This is an approximate estimate, not a guaranteed accuracy bound; route, motion, device, and initialization affect results.
Walk in. Navigate. Walk out.
Watch an independent evaluation session move from outdoors into a large store, through its interior, and back outside while maintaining a continuous track.
Trace the store aisles.
In this separate evaluation session, the entrance and exit align closely with the actual doors, while the interior path follows the observed aisle layout.
See the trail in context.
Follow a recorded path alongside terrain contours and landmarks. The evaluation app also supports downloadable USGS Topo maps for offline use.
Stay on track beyond GPS.
The recorded path remains continuous as the user moves from the street through a building complex—an environment where GPS-only positioning commonly degrades or drops out.
Find your next waypoint.
See distance and direction cues to a selected waypoint as you move, alongside positioning and altitude estimates.
Shop, check out, and exit.
This grocery store walk enters through one entrance, heads to an aisle to pick up items, turns back toward the cash register, then exits through a different entrance.
All screenshots shown are individual sessions captured on iPhone 15 Pro running iOS 26 or later, with GNSS initialization and naturally occurring degradation. They illustrate app output; map alignment is not independent ground truth and does not guarantee identical results across devices, users, or environments. The video is a separate evaluation session.
READY TO EVALUATE?
See how Rahbar performs
in your own walks.
Request a 15-day evaluation Evaluation requests are reviewed by Navstein. If the fit is right, we’ll follow up by email with next steps.