Haul Yeah

Inspiration

Drivers and dispatchers need routes that balance on-time delivery with safety. We were inspired by real-world incidents where glare, weather, and historic crash locations made otherwise-short routes riskier. Haul Yeah aims to help drivers make safer decisions — not just faster ones.

What it does

Computes optimized routes and per-segment risk scores (sunlight/glare, crash proximity, left-turn risk, weather).
Integrates with Google Maps for routing and segment instructions.
Exposes backend APIs for route optimization, sunlight risk calculation, crash-proximity analysis, and driver schedule suggestions (HOS-aware).
Presents an interactive frontend showing routes, per-segment scores, and actionable guidance.

How we built it

Backend: Flask APIs with spatial analysis (pyproj/shapely), weather enrichment, and integration with AWS Bedrock for optional AI recommendations.
Frontend: React + TypeScript + Vite; interactive map with Mapbox GL and componentized UI to surface segment-level insights.
Integrations: Google Maps for routing; optional DynamoDB persistence and AWS Bedrock for AI.

Challenges we ran into

Native spatial dependencies (GEOS, PROJ) require system libs which can block local setup.
Nationwide, high-quality accident datasets are hard to obtain and often costly.
Normalizing varied mapping provider responses required robust parsing and defensive logic.

Accomplishments we're proud of

Modular risk pipeline combining sunlight, weather, and crash proximity.
Working on Google Maps integrations and stable backend endpoints.
A frontend that clearly surfaces per-segment risk and scheduling recommendations.

What we learned

Route optimization is a complex process with various weighted factors: distance, time, speed, travel mode, historical crash data, weather, and sunlight.

What's next for Haul Yeah

Source a nationwide accident-history dataset to support long-distance delivery risk scoring.
Generate real-time optimized routes based on live risk factors (not just advice and reminders).