AI route optimization software that cut dispatch cost 42% across 180 daily loads
Logistics automation case study: Ferro Logistics replaced manual dispatch boards with AI route optimization, real-time fleet tracking, and automated carrier offers — cutting cost per load 42% in one season.
Key outcome
−42%
Lower cost per load
−42%
Cost per load
−31%
Empty miles
+18%
On-time delivery
25 hrs/wk
Dispatcher hours saved
Before
Dispatchers planned 180 daily loads in spreadsheets and radio calls. Late trucks, empty miles, and overtime burned margin every week.
After
AI-assisted dispatch with live ETA, automated carrier spot offers, and a driver mobile app — 42% lower cost per load and 31% fewer empty miles.
01 — Problem
The challenge
Ferro Logistics runs regional freight across three hubs with ~180 loads per day. Dispatch lived on whiteboards, WhatsApp threads, and a fragile spreadsheet of driver hours. Planners spent mornings rebuilding routes when a single truck delayed. Empty miles climbed. Spot-market carrier buys happened too late — rates spiked, customers waited, and ops overtime became structural.
Off-the-shelf TMS tools were either too heavy for Ferro’s size or too rigid for their mix of owned fleet and brokered capacity. They needed logistics automation: AI route optimization software that respects driver hours-of-service, dock appointments, and live traffic — plus a workflow that lets humans approve exceptions in seconds, not rebuild plans from scratch.
This engagement became Wilayah Softworks’ reference build for fleet management software with an AI decision layer — not a black-box that dispatchers ignore.
02 — Research
What we learned
We shadowed dispatch for ten days across peak and off-peak shifts, timing every decision: load assignment, relay handoffs, and spot tendering. Historical GPS and POD data from 90 days revealed that 34% of empty miles came from suboptimal return-leg pairing, not demand shortage.
We modeled constraints with operations research methods (vehicle routing with time windows) and validated against dispatcher “gut” routes. Interviews showed trust would break if the optimizer ignored dock blackout windows or preferred drivers. So explainability — why this truck, why this sequence — became a product requirement, not a nice-to-have.
Integration discovery covered ELD feeds, customer EDI appointment slots, and accounting rate tables so the AI route planner could price owned vs brokered capacity in the same view.
03 — Solution
What we built
We shipped an AI-assisted dispatch platform:
• Route optimization engine (OR-Tools) minimizing cost, empty miles, and HOS risk • Live fleet map with ETA drift alerts and dock appointment awareness • One-click exception workflows when weather or breakdowns invalidate a plan • Automated spot offers to preferred carriers when owned capacity is short • Driver mobile app for turn-by-turn jobs, delay codes, and photo POD • Dispatcher console that explains each recommendation in plain language
Cloud infrastructure on AWS scales for seasonal peaks. Redis caches live positions; PostgreSQL stores plans and audit history. The AI does not replace dispatchers — it drafts the optimal board in minutes so humans focus on customer exceptions and carrier relationships.
For shippers and 3PLs evaluating fleet management software or logistics AI automation, Ferro is the pattern: optimize continuously, keep humans in the loop, measure cost per load weekly.
Technology stack
Measurable results
Outcomes validated with the client operations team after launch — not vanity metrics.
- 42% reduction in fully loaded cost per load within one peak season
- 31% fewer empty miles across the owned fleet
- On-time delivery rate improved 18 percentage points
- Dispatchers reclaimed ~25 hours per week from manual plan rebuilds
- Spot-carrier response time fell from hours to under 12 minutes
- Fuel and overtime overrun variance dropped below finance’s target band
Frequently asked questions
Start a similar engagement
Want results like Ferro Logistics?
Tell us about your ops stack, constraints, and target metric. We will respond with a scoped approach — not a generic pitch deck.
Book a discovery call