What is Ride Service?
Ride service like OLA and UBER is a platform where rider and driver both can register and rider can avail drive service provided by driver.
Functional Requirement
- Ride Request
- A rider must be able to create a ride request by specifying a pickup (source) and drop-off (destination) location.
- Fare Estimation
- The system shall calculate and display an estimated fare for the requested trip before the rider confirms.
- Driver Onboarding & Status
- Drivers can register on the platform, publish their real-time location, and set their availability status for accepting rides.
- Request Broadcast
- After the rider confirms payment (or payment intent), the ride request shall be broadcast to all available drivers within a specified nearby radius.
- Driver Acceptance & Navigation
- A driver shall be able to accept the request and navigate to the rider’s pickup location.
- OTP Confirmation to Start Trip
- The rider can verify the correct driver by providing an OTP (one-time password); the trip shall commence only after successful OTP confirmation.
- Trip Completion & Final Fare
- On trip completion the system shall calculate the final fare and present it to the rider for payment.
- End-to-End Tracking & Monitoring
- The system shall track and log the entire lifecycle of the trip (request → acceptance → pickup → en route → completed) for monitoring, analytics, and dispute resolution.
Non Functional Requirement
- Availability : System should be highly scalable
- Scalable : As an when more drivers and riders added system should be able to scale adaptively
- Consistency : Since must be highly consistent as well
- Reliability : Fast and error free service needed
Storage & Bandwidth Estimation
India Market (December 2024)
- Uber: 840,000 daily trips
- Ola: 460,000 daily trips
Global Context (2024-2025)
- Uber: 180 million monthly active users
For design as we are planning for future, Lets assume 100 million rider and 5 million driver available on platform
Storage Estimation
- Lets assume 1 Kb of data need to save for Rider and Driver both inclusing name, id, email, driver licence no and others meta data
- Rider data = 100 Million * 1 KB = 100 GB
- Driver data = 5 Million * 1 KB = 5 GB
- Driver location metadata = 56 Byte * 5 Million = 350 Mb
- Trip Metadata = 1 Kb * 5 Million Driver * 20 trip /day = 100 GB
- Total = 210 GB * 365 = 8 TB
Bandwidth estimation
- Trip per sec = 100 Mil /86400 = 115 per second
- 115 * 1Kb * 8 = 8 MBps
- Driver req per 5 sec for location update = 1 M * 56 B *8 = 450 Kbps
Number of servers estimation
- 105 Million Req minimum /s on peak time
- Server needec – 105 M /64 K = 16 *105 = 1.6 K Server
API Design
/requestRide(riderID, lat, long, destLat, destLong, typeOfVehicle)
/showETA(rider_lat, rider_lng, destination_lat, destination_lng, ride_type)
/findNearbyDrivers(riderID, lat, long)
/confirmPickup(driverID, riderID, timestamp)
/showTripUpdates(tripID, riderID, driverID, driverlat, driverlong, time_elapsed, time_remaining)
/endTrip(tripID, riderID, driverID, time_elapsed, lat, long)
High Level Design
Building blocks we will use
Deep Dive
—
Evaluation
—
Conclusion
—
