Skip to content
SIH Buddyby Ganeev Singh
Dev
All problem statements
SIH26124Worth consideringacceptance 3/5

AI-Powered Mobile Urban Intelligence Platform Using Public Transport Fleet

Bharat Electronics Limited · Fitness & Sports · Software

Cut this to road defects plus multi-pass deduplication and say plainly what you excluded — the buses-as-sensors idea is strong, but chasing the full feature list guarantees you lose to a team that did one part properly.

What it actually is

City buses drive every major road every day and already carry cameras, but the footage is only ever reviewed after an incident. The ask is to turn the fleet into a moving sensor network that spots road damage, congestion and hazards as it drives, and feeds everything into a central city dashboard. The scope the description lists is enormous.

What to build

An onboard edge module processing bus camera feeds to detect road surface defects — potholes, damaged carriageway, waterlogging — and missing infrastructure such as absent dividers, faded zebra crossings and damaged signboards, each geotagged from the bus GPS at the moment of detection, alongside vehicle detection and counting for density estimation; plus a central platform aggregating detections from the whole fleet onto a GIS map, deduplicating repeated sightings of the same defect across multiple bus passes into a single ranked maintenance item, and generating congestion heat maps and route delay estimates for the transport authority.

Smallest thing that wins the room

Play dashcam footage of a real road and show potholes and a faded zebra crossing being detected and geotagged live, then switch to the city map where three separate bus passes over the same pothole have collapsed into one maintenance item with a confidence built from repeated sightings.

How crowded this one gets

A guess, projected from the 2025 statements — the last year where both the submission counts and the winners were published.

Moderate110–260 teams expectedroughly 1 in 95–219 wins it

Quieter than 61% of the 226 · #89 of 226 by expected field

A normal-sized field. Your idea has to be good, not miraculous.

Why: company-sponsored statements drew the smallest fields of all.

This is a guess, not a fact

Nobody has published 2026’s numbers yet. This is an analysed estimate from last year’s pattern, so please do not take it as the truth — check the live counter on the SIH portal before you decide anything. The range covers the middle half of likely outcomes, so one statement in two lands outside it. Entry closes at 500 ideas per statement, so no range goes past that — a statement that reaches the cap fills and shuts rather than drawing an unlimited crowd. The model reads only three things a team can see before choosing — software or hardware, the theme, and what kind of body posted it — and those explain about a quarter of the variation in last year’s field sizes (R² 0.25 on held-out statements). Trust the band more than the number, and the ordering more than either. It cannot see how good your idea is, which is the part that actually decides it.

The scores

The number is the shorthand. The line under it is the reason.

What you will be writing

  • YOLOv8 road damage detection (RDD2022 dataset)
  • Jetson edge inference on bus camera feeds
  • GPS-tagged detection with multi-pass deduplication
  • DeepSORT vehicle tracking and counting
  • PostGIS + Kepler.gl congestion heat maps
  • MQTT store-and-forward for intermittent connectivity
  • Urban infrastructure monitoring
  • Edge computer vision
  • Traffic analytics

Prior art to read before you start

vehicle-mounted road defect detection · crowdsourced infrastructure mapping · city-scale traffic analytics

Analysed by Claude Opus. Every score above is a judgment call with its reasoning attached — kindly cross-check this against the official statement on the SIH portal before your team commits to it.