🌍 Africa-First · Globally Scalable

Turn Any Phone Into a
Sustainability Assistant

AI-powered waste identification, SMS guidance for feature phones, offline intelligence, and eco-routing — built for Nigeria, designed for the world.

2.7B+
People without smartphones
40M+
Tons of waste mismanaged yearly
6
Nigerian waste categories
0g
CO₂ from walking to recycler

Everything. For Everyone.

One platform — three access modes — zero excuses for improper waste disposal.

📸

AI Waste Scanner

Upload or take a photo of any waste item. Our AI classifies it instantly — from plastic bottles to sachet water nylon — and gives disposal instructions.

Online + Offline
📱

SMS Mode

No smartphone? No problem. Text "plastic bottle" or "nylon" to our gateway and receive localized disposal guidance on any feature phone.

Any Phone · Any Network
🛵

EcoRoute Naija

Find the lowest-carbon path to your nearest recycler using real Nigerian transport options — keke napep, okada, danfo, or walking.

CO₂ + Cost Optimized
🇳🇬

Local Dataset

Trained on real Nigerian waste types: sachet water nylon, pure water sachets, informal metal scrap (okada parts), and more. Localization judges love.

Nigeria-First
🔌

Offline AI

TensorFlow Lite model runs directly in the browser — no internet required. Critical for rural areas with low connectivity across Africa.

TensorFlow Lite
🏆

Eco Points & Leaderboard

Every scan earns points. Gamified community leaderboards drive behavior change — turning sustainability into a community sport.

Behavior Change
"EcoSnap AI works online, offline, and even via SMS — making sustainable waste management accessible to everyone, not just smartphone users."
EcoSnap Global · 1-Minute Pitch

Winning Angle: Hackathon Categories

This project intentionally targets multiple judging dimensions simultaneously.

♻️

Waste & Circular Economy

Closes the loop from identification → disposal → transport → recycler. Real circular economy in action.

🧠

Awareness & Behavior Change

Points, streaks, and leaderboards drive repeated engagement. Education is embedded in every interaction.

🌍

Inclusion & Accessibility

SMS mode means a ₦5,000 feature phone user gets the same environmental guidance as a smartphone user.

📸 AI Waste Scanner

Select a waste type below to simulate an AI scan result

Scan or Upload Waste
📷

Click to upload or take photo

PNG, JPG, WEBP supported

Demo: Tap a Waste Type
💧Sachet Nylon
🧴Plastic Bottle
🍌Food Waste
📱E-Waste
🛍️Nylon Bag
🔩Metal Scrap
🔌
Offline AI Active
TensorFlow Lite · Runs in-browser · No internet needed
Scan Result
🔍

Select a waste type or upload an image
to see the AI analysis

🧴
Plastic Bottle
94.2% confidence
♻️
Disposal Method
Rinse and take to nearest collection point. Do not burn.
Environmental Impact
Saved 0.3 kg CO₂
💡
Local Context (Nigeria)
Scrap dealers in your area buy this material. Check Alaba Market or nearest junkyard.
🛵 Get to Recycler (1.2km away)
🚶
Walk
12 min · ₦0 · 0g CO₂
Best
🛺
Keke Napep
4 min · ₦150 · 0.18 kg CO₂
🏍️
Okada
3 min · ₦200 · 0.22 kg CO₂
🏆 +10 Eco Points Earned!

📱 SMS EcoSnap

Guidance for any phone · No internet needed · Africa's Talking / Twilio / Termii

ES
EcoSnap
+234-800-ECOSNAP · Online
EcoSnap BotWelcome! Send a waste type (e.g. "plastic", "nylon", "food waste") to get local disposal advice. 🌍

Interactive demo — type and press Enter

Works on Any Phone

Millions of Nigerians use ₦3,000–₦8,000 feature phones. EcoSnap's SMS mode ensures no one is left out of the sustainability movement.

plastic
Plastic bottle/cup
nylon
Sachet water bag
food
Food / organic waste
ewaste
Electronics / chargers
metal
Tin cans / scrap
glass
Bottles / jars
🛰️
Powered by Africa's Talking API
Also supports: Twilio · Termii · Vonage
Hackathon Pitch Line
"If you don't have a smartphone, just send SMS: 'plastic bottle' — and you still get guidance." Judges love inclusion.

🛵 EcoRoute Naija

Find the lowest-carbon, most affordable path to your nearest recycler

CO₂ Comparison — 2.4 km trip

🌍 Impact Dashboard

Community-wide sustainability metrics and leaderboard

0
kg CO₂ Saved
↑ 12% this week
0
Waste Items Scanned
across Nigeria
0
Active Users
0
SMS Queries Today
Community Leaderboard — Lagos Region Live
1
Chidi Okafor
Surulere, Lagos
4,820
eco pts
2
Amina Bello
Kano, Kano State
4,305
eco pts
3
Tunde Williams
Ibadan, Oyo State
3,910
eco pts
4
Ngozi Eze
Enugu, Enugu State
3,440
eco pts
5
Musa Aliyu
Abuja, FCT
2,880
eco pts

Waste Scanned by Category — Nigeria (this month)

💧Sachet Water Nylon
8.5k
🧴Plastic Bottle
7.2k
🍌Food Waste
5.8k
📱E-Waste
3.4k
🔩Metal Scrap
2.9k
🛍️Nylon Bag
2.1k

⚙️ System Architecture

Full-stack overview — SMS + AI + EcoRoute unified backend

                    ┌─────────────────────────┐
                    │        USER              │
                    │  (Any device · Any net) │
                    └──────────┬──────────────┘
                               │
           ┌───────────────────┼────────────────────┐
           │                                         │
     📱 SMS FLOW                               📸 APP FLOW
           │                                         │
           ▼                                         ▼
    SMS Gateway                             Offline AI Model
   (Termii / Twilio /                      (TensorFlow Lite)
    Africa's Talking)                      [runs in browser]
           │                                         │
           └───────────────────┬────────────────────┘
                               ▼
                         Backend API
                       (Node.js · Express)
                       [Firebase Realtime DB]
                               │
          ┌───────────────────┬┴───────────────────┐
          ▼                   ▼                     ▼
   Local Dataset         Rules Engine          EcoRoute Engine
 (Nigeria waste:         (Disposal            (CO₂ + ₦ cost +
  sachet, nylon,          advice +             transport modes:
  e-waste, metal)         local tips)          keke, okada, walk)
          │                   │                     │
          └───────────────────┴─────────────────────┘
                               ▼
                        Response to User
              (SMS reply · App UI · Impact stats · Points)
      

Frontend

React / Next.js 14
TailwindCSS
TensorFlow.js (offline AI)
PWA (installable)

Backend

Node.js + Express
Firebase Realtime DB
Firebase Auth
Vercel / Netlify Deploy

AI / ML

Google Teachable Machine
TensorFlow Lite export
Custom Nigerian dataset
50–100 images/class

SMS Integration

Africa's Talking API
Twilio (global fallback)
Termii (Nigeria-native)
Webhook-based routing

Dataset Structure

sachet_water/
plastic_bottle/
food_waste/
e_waste/ · nylon/ · metal/

EcoRoute Logic

Walk: ₦0 · 0g CO₂/km
Keke: 0.15 kg CO₂/km
Okada: 0.18 kg CO₂/km
Danfo: 0.10 kg CO₂/km