About the Project
Platanos is a hackathon project built to improve last-mile logistics for Damm/DDI delivery operations. The idea was inspired by a real operational challenge: delivery drivers do not only need an efficient route, they also need the truck to be loaded in a way that makes every stop faster and easier.
In traditional delivery planning, routing and truck loading are often treated as separate problems. However, in real distribution work, both decisions are connected. If the first customers are loaded in an inaccessible part of the truck, the driver loses time searching, moving products, and reorganizing the load during the route.
Our project connects these two dimensions:
- route planning,
- truck loading,
- customer time windows,
- delivery order,
- product type,
- truck capacity,
- and reverse logistics.
What We Built
We built a prototype web platform for Damm delivery drivers. After logging in, a driver can see assigned routes, estimated duration, customer availability, truck loading information, and operational recommendations.
The demo focuses on three real sample routes:
- DR0006 — Mollet
- DR0017 — Mollet / Granollers
- DR0038 — Granollers / Canovelles
For each route, the platform shows:
- number of customers,
- number of order lines,
- total pieces,
- estimated route duration,
- customer time windows,
- truck loading plan,
- 2D truck layout,
- operational alerts,
- and recommended loading strategy.
How It Works
The solution combines two main modules.
1. Route Planning
The route module estimates the best delivery sequence using:
- customer location,
- estimated travel time,
- time windows,
- service time,
- and route constraints.
We also explored the idea of grouping nearby establishments into a single truck stop when several businesses are close enough to be served from one parking point.
2. Truck Loading
The loading module assigns products to truck positions based on the expected delivery order. The goal is to make the first stops easier to access and avoid unnecessary product movement during the route.
The truck is represented as a simplified loading layout with:
- 6 pallets,
- 12 rows,
- 6 columns,
- multiple height levels,
- and product categories such as boxes, barrels, returnables, and other units.
What We Learned
During the project, we learned that the best logistics solution is not always the one with the shortest route. A good delivery plan must balance several objectives:
- minimizing driving time,
- reducing unloading effort,
- respecting customer availability,
- improving truck accessibility,
- managing returnables,
- and keeping the loading process realistic for the warehouse team.
We also learned that data quality is critical. Some addresses are difficult to geocode automatically, so the system must clearly distinguish between verified data and data that requires manual review.
Challenges
One of the main challenges was connecting route optimization with physical truck loading. It is easy to optimize a route on a map, but much harder to make that route practical for the driver once the truck is full.
Another challenge was building a demo that stayed realistic. We avoided inventing operational results and focused on making assumptions visible. When exact data was not available, we treated it as a limitation rather than hiding it.
We also had to simplify some real-world constraints, such as traffic, parking availability, exact unloading time, and detailed stacking rules. These would be important additions in a production pilot.
Why It Matters
This project matters because it focuses on the driver’s real workflow. The driver does not only need to know where to go next; the driver needs to know where the product is, how to access it quickly, and how the route will affect the loading and unloading process.
By combining route planning and truck loading, the system can help reduce:
- time spent searching for products,
- unnecessary movements inside the truck,
- failed deliveries due to timing issues,
- loading inefficiencies,
- and operational uncertainty.
Future Improvements
The next step would be to connect the prototype with real validated customer coordinates, historical delivery times, live traffic data, and warehouse preparation systems.
Future versions could include:
- real GPS-based routing,
- validated customer master data,
- driver feedback,
- automatic PDF loading sheets,
- integration with SAP or DDI systems,
- better support for returnables,
- and a pilot test with real delivery teams in Mollet.
Final Vision
Our vision is simple:
Do not optimize only the route. Optimize the full delivery operation.
Platanos helps connect warehouse preparation, truck loading, route planning, and driver execution into one operational decision.
Log in or sign up for Devpost to join the conversation.