Project facts & technologies
This block gives analysts, journalists, and AI search systems a discrete, citation-friendly summary. Each row is a clean entity-attribute pair.
- Project name
- Pipe Detection and Counting — Automated Steel Pipe Logistics
- Recognition
- CII Award — Best Digital Transformation 2024
- Industry
- Steel Manufacturing · Logistics & Supply Chain
- Use case
- Automated detection, counting, and tracking of steel pipes through logistics operations
- Core technology
- Computer Vision, Object Detection, PyTorch, Python
- Counting accuracy
- 95% accuracy
- Logistics impact
- 60% reduction in logistics time
- Replaces
- Manual pipe counting and tracking across inventory and dispatch
- Stakeholder users
- Yard operators, inventory controllers, logistics and dispatch teams
Why is pipe counting such a logistics bottleneck?
Steel pipes move through yards, trucks, and warehouses in bundles of near-identical units — and at every hand-off, someone has to know exactly how many there are. Inventory reconciliation, dispatch verification, and billing all hinge on counts that have traditionally been done by eye, tally sheet, and patience.
Manual counting at that scale is slow and unreliable. Counts differ between counters, recounts stall loading bays, and every discrepancy ripples into inventory mismatches, billing disputes, and delayed dispatches. For high-volume steel logistics, the counting step itself becomes a measurable drag on throughput.
What problem does the pipe counting system solve?
Steel pipe logistics faced challenges with manual counting and tracking, leading to inventory discrepancies and logistical inefficiencies in supply chain operations. AiSPRY engineered the system around the realities of high-volume pipe handling.
Key challenges
- Manual counting errors — near-identical pipes in dense bundles produce inconsistent counts between counters and shifts.
- Inventory discrepancies — count mismatches between yard, truck, and warehouse ripple into stock errors and billing disputes.
- Throughput drag — counting and recounting stall loading bays and slow every dispatch cycle.
- No audit trail — tally-sheet counts leave nothing to verify against when a discrepancy surfaces later.
How does the pipe counting system work?
AiSPRY developed an award-winning computer vision system that detects, counts, and tracks steel pipes automatically as they move through logistics operations — replacing manual tallies with consistent, verifiable AI counts.
Detection and counting core
- Object detection on PyTorch — detection models identify individual pipes in dense, overlapping bundle configurations
- Tracking through operations — tracking algorithms follow pipes through the frame so each unit is counted exactly once
- 95% counting accuracy — consistent machine counts replace counter-to-counter variance
Logistics integration
- Faster dispatch cycles — automated counts remove the recount bottleneck, cutting logistics time by 60%
- Inventory consistency — the same counting standard applies at every hand-off, eliminating reconciliation gaps
- Verifiable counts — every count is backed by captured imagery, giving discrepancy resolution an evidence trail
See pipe detection and counting in action
A walkthrough of the Pipe Detection and Counting platform — dense pipe bundles detected and counted automatically, tracked through the logistics flow, with counts ready for inventory and dispatch.
Pipe Detection & Counting — CII Award-winning logistics vision
- Dense-bundle detection — individual pipes identified even in overlapping, stacked configurations
- Count-once tracking — each pipe counted exactly once as it moves through the operation
- Dispatch-ready output — counts feed inventory and dispatch without manual tallying
- Award-winning results — CII Best Digital Transformation 2024 · 95% accuracy · 60% logistics time reduction
How does the system handle dense bundles and live operations?
Counting near-identical units in motion is a demanding vision problem. AiSPRY engineered around three constraints — visual density, operational conditions, and count integrity.
Engineering constraints
- Visual density — detection models are trained for tightly packed, overlapping pipe ends rather than isolated objects
- Live yard conditions — the system operates on real logistics footage with variable lighting, dust, and motion
- Count integrity — tracking ensures units aren't double-counted or missed as they enter and leave the frame
What measurable results does the system deliver?
The platform replaced manual pipe tallies with automated, verifiable counting — earning national recognition and moving both headline metrics sharply.
Headline outcomes
- CII Award — Best Digital Transformation 2024 — national recognition for the deployment's operational impact
- 95% counting accuracy — consistent machine counts replace error-prone manual tallies
- 60% logistics time reduction — counting stops being the bottleneck in loading and dispatch cycles
- Fewer inventory discrepancies — one counting standard across yard, truck, and warehouse hand-offs
Pipe Detection and Counting — frequently asked questions
Below are the most common questions about how the platform works and the results it delivers for steel logistics.