Inspiration

WasteWatchers was inspired by the amount of food that becomes waste while being transported in refrigerated trucks. When improper temperature control or handling puts a shipment at risk, logistics teams may spend valuable time checking separate systems, calling drivers, and deciding whether the load can still be recovered. During that delay, product quality continues to decline.

We wanted to build a tool that helps shipping managers understand shipment health quickly and act before recoverable food becomes waste.

What it does

WasteWatchers is a monitoring and decision-support platform for refrigerated agricultural shipments.

It receives truck temperature telemetry, compares it against the safe temperature range for the commodity, estimates remaining shelf life, and classifies each shipment as healthy, watch, at risk, or critical.

The dashboard helps shipping managers:

Monitor shipment conditions
Identify pallets that are in danger
View current and historical temperatures
Compare temperatures against the safe range
Track remaining shelf life
See the estimated product value that can still be protected
Receive rerouting recommendations
See how much time remains to act
Approve a reroute
Reject a shipment
Send a shipment for manual review
Simulate cooling failures, temperature spikes, telemetry outages, and recovery

The goal is to make shipment health easy to understand without requiring the user to interpret a large amount of technical data.

How we built it

We built WasteWatchers with:

Python 3.12
FastAPI
Pydantic
SQLite
Next.js
Tailwind CSS
pytest

The FastAPI backend handles telemetry ingestion, commodity thermal profiles, shelf-life estimation, shipment risk classification, rerouting recommendations, simulation state, manager decisions, and persistence.

The Next.js frontend presents this information as a logistics control center. It includes a prioritized shipment queue, refrigerated trailer visualization, pallet condition indicators, temperature and shelf-life summaries, simulation controls, and decision panels.

We also used GitHub for collaboration and followed a spec-driven workflow using Spec Kit and Codex. We defined the feature requirements, created an implementation plan, divided the work into tasks, and tested each major phase before moving forward.

Challenges we ran into

One major challenge was combining work that had been developed in separate repositories. We created an integration branch, added the shared repository as a remote, cherry-picked the backend implementation, resolved conflicts, and merged the work through a pull request.

We also faced challenges with:

Connecting the Next.js frontend to FastAPI
Configuring API URLs and browser access
Managing simulation state
Preventing overlapping simulation requests
Preserving the selected shipment during live updates
Designing a trailer visualization that clearly showed pallet risk
Making the dashboard responsive across different screen sizes
Keeping business logic in the backend instead of duplicating it in the frontend
Translating technical measurements into language that a shipping manager can understand quickly

Accomplishments that we’re proud of

We are proud that WasteWatchers developed from a basic monitoring concept into a working decision-support platform.

Our main accomplishments include:

Building a documented FastAPI backend
Creating seeded demo scenarios
Supporting healthy, watch, at-risk, and critical shipment states
Building a live telemetry simulation
Creating a refrigerated trailer and pallet risk visualization
Estimating remaining shelf life
Generating salvage rerouting recommendations
Supporting manager decisions
Maintaining a passing automated test suite
Successfully integrating independently developed frontend and backend work

We are especially proud that the dashboard is designed for fast visual understanding. A shipping manager does not need to be a data scientist to see which load needs attention and what action should be taken.

What we learned

We learned that reducing food waste in transportation is not only a data problem. It is also a communication and decision-speed problem.

Temperature readings are useful, but they do not solve the problem by themselves. Managers need to quickly answer four questions:

What is wrong?
How serious is it?
How much time remains?
What should be done next?

We also learned more about API design, frontend and backend integration, state management, simulation systems, SQLite persistence, automated testing, responsive interface design, Git collaboration, and cloud deployment.

Most importantly, we learned how important it is to translate technical data into clear operational guidance.

What’s next for WasteWatchers

The next step is to connect WasteWatchers to real refrigerated fleet telemetry and operational data.

Future improvements include:

Live truck and trailer sensor integration
More commodity thermal profiles
Route and traffic-aware rerouting
Real-time cold storage availability
Driver and dispatcher alerts
Predictive spoilage models
Historical shipment analysis
Mobile access
Authentication and role-based permissions
Persistent cloud database storage
Fleet-management platform integrations
Support for more carriers and distribution networks

Our long-term goal is to help logistics teams recover more agricultural shipments, protect product value, and prevent food from becoming waste during transportation.