HVAC Manufacturing
Reduction in empty delivery miles
Annual logistics spend reduction
About
The manufacturer wanted to slash GCC delivery costs for bulky HVAC ducting while staying fully compliant with axle rules and driver hours. Manual routing left trucks 17% empty and 11% late, and axle-load breaches attracted fines and regulatory flags for hours-of-service violations. We built a branch-and-price routing engine modelling three fleet types against live traffic data.
Industry
HVAC Manufacturing
Company size
500 – 1,000 employees
Founded
1988
The Company
Bulky freight, tight regulation, thin margins
The business manufactures HVAC ducting and delivers it across the GCC, where bulky, low-density freight makes truck utilisation naturally difficult and regional axle-load and hours-of-service regulations add real compliance risk to every route.
Delivery planning had been handled manually, with dispatchers building routes from experience rather than from a model that could simultaneously respect fleet-type constraints, axle patterns, and live traffic conditions.
The challenge
Manual routing left a fifth of every truck empty
Manual routing left trucks running 17% empty on average and 11% of deliveries arriving late. Axle-load breaches on badly sequenced routes were attracting fines, and hours-of-service violations were repeat enough that regulators had begun flagging the fleet.
None of this was a driver problem — it was a planning problem. Dispatchers had no systematic way to weigh fleet-type capacity, axle limits, and live traffic together when building a route, so they defaulted to familiar patterns that didn't account for any of it precisely.
The Solution
A branch-and-price VRP engine modelling three fleet types against live traffic
We built a branch-and-price vehicle routing engine modelling three distinct fleet types, their axle patterns, and live HERE-traffic speeds simultaneously. Roughly 17 million daily GPS pings flow into Snowflake and drive a Power BI map that highlights routing inefficiency by 5 km grid cell.
Dispatchers now build routes against a model that enforces axle-load and hours-of-service constraints automatically, rather than relying on a human to remember every regulatory limit route by route.
The Results
19% fewer empty miles, zero HOS violations
Empty miles fell 19%, and annual logistics spend dropped by Rs. 3.1 crore as routes were built against actual fleet capacity and traffic conditions rather than dispatcher habit. Zero hours-of-service violations were recorded in the quarter following rollout.
The Power BI grid-level map also gave the logistics team, for the first time, a way to see which specific corridors were structurally inefficient — informing depot and fleet-mix decisions beyond day-to-day routing.
KEEP READING
Explore more studies on AI, analytics, and enterprise intelligence.
