Runyenyesa/yitayo

Yitayo — Powering Uganda's Electric Journey. An AI-powered real-time transit platform built for E-Bus Xpress Kayoola electric buses. Features live GPS tracking, passenger ETAs, AI demand prediction, fleet management and a public PWA — scaling from 8 buses in Kampala to 1,500 buses across 14 Ugandan cities by 2031. # Yitayo - E-Transit Infrastructure Tracker Yitayo is a high-utility, open-source transit tracking system designed for the **Kayoola EVS (Electric Vehicle Series)** bus fleet in the Kampala Metropolitan Area [KMC]. The defining core constraint of this project is **Zero-Hardware tracking**. Traditional GPS tracking boxes are expensive, prone to vehicle circuit overloads, and vulnerable to theft or maintenance tampering. Yitayo moves the tracking workload entirely into software by utilizing a distributed **Citizen-Sensor Network** and static operational infrastructure handshakes. --- ## How the Tracking System Functions Yitayo divides its tracking lifecycle into two distinct execution paths to achieve continuous remote asset location logs without active onboard hardware trackers: Use code with caution.[DAYTIME TRACKING]Commuter App Checks In ➔ Mobile GPS Aggregation ➔ Weighted Vector Filtering ➔ Real-Time Map Render[NIGHTTIME LOCKDOWN]Driver Logs Off Shift ➔ Depot QR Authentication ➔ KMC Charging Dock Network Handshake ➔ Status Lock ### 1. Daytime Operations: Crowdsourced Active Loop Instead of tracking the *machine*, Yitayo tracks the *trip* using the devices inside the vehicle space. * **The Scan Check-In:** High-durability vinyl sheets printed with location-unique, cryptographic QR codes are affixed across the interior walls of each Kayoola bus. * **The Token Anchor:** When a passenger or conductor launches the application and validates their ride via the QR code, the platform requests brief, low-power background location coordinate telemetry. * **Deterministic Algorithmic Matching:** The backend ingests these individual passenger device coordinate streams. It applies a rolling time-window filter, drops erratic outliers (e.g., someone walking away from a window), and calculates a weighted velocity vector average. If 15 passenger devices move down Jinja Road simultaneously at 40 km/h, the server mathematically verifies the bus's precise real-time placement ma …