Inspiration
Every minute wasted in delivery translates to higher costs, traffic congestion, and CO₂ emissions. Existing routing solutions like Dijkstra/A* or even Google OR-Tools don’t adapt well to real-time traffic and multi-vehicle logistics. We wanted to build an algorithm-first solution that’s faster, smarter, and greener.
What it does
D-RLO is an AI + algorithm-driven logistics optimizer that dynamically finds the best delivery and ride-sharing routes in real time.
Models cities as a weighted graph (roads, distances, time).
Uses real-time traffic predictions (via ML regression/LSTM).
Optimally assigns drivers/vehicles using Hungarian Algorithm / Min-Cost Max-Flow.
Adds a genetic algorithm layer for minimizing fuel/EV battery usage.
Provides a dashboard for fleet operators + API for developers.
How we built it
Frontend/UI: Stunning landing page + dashboard (Apple-inspired, Dribbble references).
Backend: Node.js + Express + Python services for algorithms.
Algorithms: Modified Dijkstra/A*, Hungarian Algorithm, GA-based optimization.
ML Layer: Traffic prediction using time-series datasets.
Database: PostgreSQL / Firebase for real-time request handling.
Hosting: Deployed on Vercel + Cloud backend for scale.
Challenges we ran into
Balancing accuracy vs. speed in route optimization.
Handling large-scale graphs for metro-level logistics.
Integrating multiple algorithms (graph + ML + GA) seamlessly.
Building a real-time dashboard that updates instantly.
Accomplishments that we're proud of
Created an algorithm-first logistics engine that beats vanilla Dijkstra/A* in benchmarks.
Integrated ML + optimization + graph theory into one cohesive product.
Designed a stunning, investor-ready UI/UX that makes complex algorithms feel simple.
Built a working demo that scales from small towns to mega-metros.
What we learned
How to blend algorithms and AI for real-world optimization.
Trade-offs between runtime complexity and real-time decision-making.
Building developer-friendly APIs for algorithm-heavy backends.
Importance of clean UI/UX in making technical products approachable.
What's next for D-RLO (Dynamic Real-Time Logistics Optimizer)
Scaling to drones & EV charging logistics.
Adding multi-objective optimization (cost, time, emissions).
Partnering with delivery, ride-sharing, and EV fleet companies.
Open-sourcing the core algorithm engine for developers & researchers.
Log in or sign up for Devpost to join the conversation.