Soft Suave - Live Chat 
💬 Hi there! Need help?
1
Soft Suave - Live Chat
1
The client operates traffic-monitoring infrastructure along a busy Mexican highway corridor and set out to build a real-time vehicle axle counting solution. The goal: efficiently count axles, track vehicles by lane and measure throughput in real time — with special attention to lifted or floating axles — while optimising queue counting and top-view processing. And the system had to hold that accuracy in dense, congested traffic, where vehicles overlap in the top-view camera feed.
Real traffic is unpredictable — vehicles overlap, configurations vary, and conditions change every second. Building a reliable intelligence system required solving these challenges to deliver accurate, real-time insights across high-density road environments.
Commercial vehicles raise axles clear of the road — traditional counting systems frequently miscounted these lifted axles or missed them entirely.
Real-world traffic mixes many vehicle classes and trailer setups that strain rule-based counting and demand a robust AI approach.
Queue detection and classification had to run across several lanes simultaneously at near real-time speed to be useful.
In dense, congested traffic, vehicles overlap in the top-view camera feed, making clean per-vehicle axle separation far more difficult.
Soft Suave built the system as a single computer-vision pipeline over top-view camera feeds—real-time detection, lane-aware tracking, specialised axle recognition, and queue analytics in one platform, deployed in production across multiple high-traffic sites and tuned to maintain accuracy in dense, congested traffic.
Advanced computer-vision algorithms process top-view feeds to count vehicle axles with high precision, even in dense traffic — with a specialised module identifying lifted or floating axles for complete classification accuracy.
A sophisticated tracker maintains each vehicle's identity across frames and ties it to its lane, delivering accurate per-lane throughput measurement in real time.
Optimised queue-counting algorithms detect congestion and vehicle accumulation as they build, giving operators the insight needed for proactive traffic responses.
"Every axle counted, every lane tracked, every queue seen early — one production vision pipeline turning live camera feeds into traffic intelligence."
The solution integrates advanced computer vision models, multi-lane tracking, and real-time analytics into a unified platform. These capabilities enable accurate vehicle classification, axle counting, congestion monitoring, and operational insights across challenging traffic scenarios.
Transforming complex traffic data into actionable intelligence, the platform achieved significant improvements in accuracy, speed, and reliability — enabling smarter traffic operations at scale.
Enhanced detection accuracy over the previous monitoring baseline
Improvement in vehicle tracking and traffic decision-making
Better handling of complex vehicle and trailer configurations
Faster access to real-time traffic data
The client needed more than detection—a real-time vision pipeline handling lifted and floating axles, dense multi-lane traffic, and live queue analytics, deployed across multiple high-traffic locations. Soft Suave delivered a scalable, production-ready platform with proven ROI and a future-proof architecture - accuracy that holds in the densest congested traffic, with every lane, vehicle, and axle accounted for in real time.
Download this practical case study to learn how our Vision AI platform delivers real-time axle counting, lane tracking, and congestion monitoring across live highway traffic.
Explore how we help startups and SMBs grow, launch successful products, and solve challenges with tech and strategy—backed by real results in every case study.
Book Free Consultation
Get a 30-minute free consultation from a field expert. Validate your idea for free and get a rough quote once you complete this form.
Madhu Kadiyala
Technology Consultant
