Trucker AI

Inspiration

We realized that dispatching a trucking fleet involves way too much guesswork. Dispatchers spend half their day calling drivers just to check locations and driving hours. Once a truck is on the highway, it goes dark, and after delivery, the team spends days chasing down messy paper receipts just to send an invoice. We wanted to build a single smart tool to take that mental load off dispatchers and drivers alike.

What it does

Our project is an AI-Native Fleet Operations Assistant that tackles three headaches:

Smart Dispatch: Our AI matching algorithm instantly ranks the best drivers for a new load based on distance, weather, and remaining driving hours.
Proactive Safety Alerts: It automatically warns dispatch if a truck has been suspiciously idle or suddenly decelerates. Drivers also get a one-tap SOS button for emergencies.
Automated Billing: Drivers snap a photo of their delivery documents, and our backend instantly generates a clean invoice and emails it to the client.

How we built it

We used React and Vite (styled with Tailwind CSS) for a highly responsive, user-friendly frontend, integrating React-Leaflet for live mapping. On the backend, we used a Node.js/Express server hooked up to a SQLite database via the Prisma ORM. We wrote custom background scripts to power the AI matching algorithm and the invoice generator.

Challenges we ran into

Getting the matching algorithm right was tricky! We had to balance multiple variables so that the system wouldn't just pick the closest driver, but rather the closest driver who actually had enough legal driving hours left to finish the job. Also, mocking live truck telemetry (like speed drops or idle times) in a static hackathon environment took some creative problem-solving.

Accomplishments that we're proud of

Bridging the gap between a complex dispatcher dashboard and a super-simple driver mobile app in one weekend felt amazing. Watching our automated billing pipeline work, where a simple file upload instantly spits out a finished invoice, was definitely a massive "high-five" moment for the team.

What we learned

We learned a ton about the strict regulatory realities of the trucking industry, especially Hours of Service compliance. Technically, we got a crash course in managing real-time geospatial state in React without causing the frontend to lag.

What's next for Trucker AI

We want to integrate our system directly with physical Electronic Logging Devices (ELDs) to pull real-time location and speed data. We’re also planning to add predictive maintenance to our AI algorithm (avoiding assigning long hauls to trucks due for repairs) and implement OCR so the system can automatically read and extract text straight from the drivers' receipt photos.

Presentation: https://docs.google.com/presentation/d/1ProaZhzXIycwaGaRMWeGV0N6Rn1m-Bb1qpRLf6ha2Y0/edit?usp=sharing