How We Save Couriers’ Time: Not Just a Route—Intelligent Optimization

When you enter a list of addresses into a standard navigation app, it simply connects the dots in the order you provided. But true route optimization begins when the system decides the best order for you, while fully accounting for the real road network, vehicle type, time windows, and dozens of other practical constraints. That’s exactly how our service works—and why couriers get routes that genuinely save time, fuel, and effort.

Three Stages That Turn Chaos Into a Clear Plan

Our approach is built on rigorous, multi-stage optimization. We don’t just “draw lines on a map”—we solve a complex engineering problem by breaking it down into three logical phases:

Stage 1. Building an Accurate Time and Distance Matrix

The first step is to construct a complete matrix between every pair of locations in the route: from depot to each delivery point and back.
But this isn’t straight-line distance. It’s realistic travel time, calculated while respecting:

• Vehicle type (pedestrian, bicycle, car, van, truck, etc.),
• Road restrictions (one-way streets, bridge height/weight limits, prohibited turns),
• Toll roads (included or excluded based on user preferences),
• Live and historical traffic patterns,
• Legal access rules (e.g., low-emission zones, pedestrian-only areas).

To compute this efficiently, we use an optimized **bidirectional A*** algorithm on a high-fidelity road graph. This enables fast, precise calculation of actual travel time between any two points, even in massive metropolitan areas with millions of road segments.

The result? A realistic, context-aware matrix—the foundation for intelligent sequencing.

Stage 2. Optimizing Visit Order with advanced Linear Programming technics

Once the matrix is ready, mathematical optimization takes over.
We formulate the visit sequencing problem as an Linear Programming (LP) model, where:

• The objective is to minimize total time (or distance),
• Constraints include time windows, vehicle capacity, priority rules, forbidden stop combinations, and business-specific policies.

This approach finds a globally optimal—or near-optimal—visit sequence, something greedy heuristics like “nearest neighbor” can’t achieve, especially in complex scenarios (e.g., morning vs. evening deliveries, mixed pickup/drop-off tasks).

Stage 3. Generating the Final Navigation Path

After the optimal sequence is determined, we re-run precise pathfinding between each consecutive pair of stops—this time using the exact order from Stage 2.
This ensures the final route:

• Fully complies with road rules for the selected vehicle,
• Avoids prohibited or unwanted segments (e.g., tolls, unpaved roads),
• Uses the most efficient turns and transitions,
• Renders as a single, smooth, actionable navigation path on the map.

Why Does This Matter in Practice?

Because the real world is full of constraints.

• Skipping a toll road might add 15 minutes—but save $10.
• Visiting stops in the wrong “convenient” order can force a courier to crisscross the same neighborhood three times.
• A delivery van can’t enter a narrow alley that’s fine for a bicycle.

Our system accounts for all these details at every stage, turning route planning from guesswork into a precise, repeatable process.

The Result: Fewer Kilometers, More Deliveries, Higher Profit

Users of our service consistently report:

• 25–40% reduction in total driving time,
• 20–30% more deliveries completed per shift,
• Lower stress for couriers—routes feel logical, not chaotic.

Optimization That Understands Context

We don’t just build routes.
We build smart delivery plans, tailored to your vehicle, your city, and your business rules—all in seconds, with no complex setup.

Try it yourself—and see how the right algorithms make all the difference.