Fleet Fuel Optimization Without the Black Box: A Technical Look at How OptiMile Pro Actually Works

No mystery algorithms. No painful integrations. No vendor lock-in. Here's the architecture, the math, and the security model, so you can evaluate it properly.

You've seen the pitch deck. Someone from ops or the fleet owner forwarded you a link about fuel savings (27-34%) and asked you to "take a look at this."

So here you are. And your first three questions are probably:

1. How does it actually work?
2. What does it need to integrate with?
3. What's the security and data exposure?

Fair. Let's answer all three with enough technical detail to actually be useful.

The Problem Statement (From an Engineering Perspective)

Trucking fleets negotiate contracted fuel pricing with major networks: Love's, Pilot Flying J, TA/Petro, and others. These contracts provide per-station pricing that varies significantly across locations, even within the same network.

The optimization problem: given a fixed route, a set of contracted station prices along that route, fuel tank constraints, and a fuel network topology, determine the optimal sequence of fuel stops and fill quantities that minimizes total fuel cost.

This is a constrained optimization problem with the following characteristics:

• Decision variables: which stations to stop at and how many gallons to purchase at each
• Constraints: tank capacity, minimum fuel thresholds, station availability within a defined corridor, driver MPG, fuel on board at final destination, and contracted network membership
• Objective function: minimize total fuel expenditure across the route

It's not a trivial problem. For even a 5-stop route with dozens of candidate stations per stop, the solution space is combinatorially large. Spreadsheets and manual heuristics leave significant money on the table, which is exactly what the route data shows.

How OptiMile Pro Solves It

Input: Contracted Pricing Ingestion

The system starts with your fleet's actual contracted pricing. Not retail. Not national averages. Your rates, at your stations, under your agreements.

Pricing data is ingested via:

• Direct feed from fuel networks: where available, we pull contracted pricing programmatically
• Fleet-provided rate tables: CSV/Excel upload of your negotiated rates by station or network

The pricing model accounts for:

• Station-level price variation within the same network
• Time-based price fluctuations
• Volume-based pricing tiers where applicable
• Multi-network optimization (if your fleet has contracts with multiple networks, the system optimizes across all of them)

Processing: Route Corridor Analysis

Given a route (origin, destination, waypoints), the system defines a corridor (typically 25 miles on either side of the planned route) and identifies every contracted fuel station within that corridor.

Each candidate station is geocoded and mapped to the route with:

• Distance from route centerline
• Position along the route (mile marker equivalent)
• Detour cost (time and fuel to reach the station and return to route)
• Station attributes (amenities, parking, operational hours)

This generates a directed graph of possible fueling sequences along the route, with stations as nodes and route segments as edges.

Optimization: Dynamic Programming Engine

The solver evaluates the cost of every feasible fueling sequence, accounting for:

• Purchase price per gallon at each station (contracted rate)
• Fuel consumed between stations (route distance x fleet MPG)
• Detour fuel cost for off-route stations
• Tank constraints (capacity, current level, minimum reserve, fuel levels at arrival)

Output: an ordered list of fuel stops with specific gallon quantities at each, total route fuel cost, and the delta versus unoptimized (baseline) fueling.

What "Unoptimized" Means

The baseline comparison isn't arbitrary. We model unoptimized fueling as the typical driver behavior: fueling at the most convenient contracted station when the tank reaches a threshold (usually 1/8 tank). This reflects real-world patterns, not a worst-case strawman.

These are real routes with real contracted pricing, not synthetic benchmarks.

Integration Architecture: Start With One, Scale When Ready

This is where most fleet tech tools fail your evaluation. They require TMS integration before you can see a result. Six weeks of API work before anyone knows if the product delivers value.

OptiMile Pro is designed for zero-integration onboarding. You get more out of the system if you integrate with your telematics provider but you sure don't have to!

The integration path is incremental. You don't need to boil the ocean. Start manual, automate what makes sense, expand at your own pace.

Security and Data Handling

Data Architecture

Compliance

• No third-party data sharing: your data is used exclusively to generate your fuel plans. Period. We don't sell data, share it with fuel networks, or use it to train models that benefit other customers.
• Access controls: role-based access with audit logging. You control who in your organization sees what.

Infrastructure

• Cloud-hosted on major provider infrastructure (GCP)
• Multi-AZ deployment for availability
• Automated backups with point-in-time recovery
• Infrastructure as code: reproducible, auditable deployments

What We Don't Need Access To

This is as important as what we do need. OptiMile Pro does not require:

• Access to your TMS (unless you want integration)
• Driver PII
• Load or customer data
• Financial systems
• ELD/HOS data

Vendor Evaluation Checklist

If you're doing a proper vendor evaluation, here's what you'd want to verify. We'll provide documentation or live demonstration for each:

What a Proof of Concept Looks Like

We're engineers talking to engineers here. A POC should answer one question: does this work with our data, on our routes, at our scale?

Here's the POC structure:

The entire evaluation requires zero changes to your existing infrastructure. No staging environment. No sandbox integration. No IT tickets. Just data in, results out.

Request a Technical Walkthrough

If you've read this far, you're the person who's going to make the recommendation. You need to see the system, ask hard questions, and verify the claims.

We'll show you:

• The optimization engine running on sample routes (or yours, if you bring them)
• A live proof of concept plan using your actual contracted pricing

No sales team. You'll talk to the people who built it. Bring your hardest questions.