Back to WorkCase Study 02
IoT / Mobile

SAHEEBA

Real-time IoT & Community Platform combining mapping precision with massive concurrent users.

Role

Mobile Engineering

Platform

iOS & Android

Timeline

16 Weeks

Core Stack

React Native, WebSockets

The Challenge

Creating a real-time tracking application is rarely a UI problem—it is a battery drain and data throughput problem.

Saheeba needed a cross-platform mobile app capable of broadcasting continuous high-frequency GPS telemetry to a central server, while simultaneously receiving and rendering the live positions of hundreds of other community members on a dynamic interactive map.

Standard polling methods exhausted phone batteries in under two hours and resulted in choppy, inaccurate "jumping" cars on the map. We needed a deeply integrated native solution.

The Architecture

We abandoned HTTP polling in favor of a persistent bi-directional WebSocket connection paired with custom native iOS/Android location modules.

Smart Telemetry Throttling

We wrote native OS background tasks utilizing accelerometer data to determine if a vehicle was stationary, dynamically adjusting the GPS ping rate from 1s to 30s to save battery.

Vector Mapping Interpolation

Instead of raw coordinate jumps, we used Mapbox GL with custom mathematical interpolation to animate vehicle markers smoothly along recorded road splines.

Engineering Impact

The result was a highly performant application that felt like magic—real-time physics mixed with a community social layer.

<50ms

Latency

Glass-to-glass websocket transmission time.

-65%

Battery Usage

Reduction in background power consumption.

60fps

Map Rendering

Sustained frame rate with 500+ markers.

Ready to scale your own architecture?