AI-Powered Weld Intelligence

Weld defects caught before the inspector arrives

WarpSense is a clip-on sensor system that monitors five structural parameters in real time during the weld. A pass-or-flag verdict is delivered the moment the joint is finished.

3
Active production shipyards
4
Named industry leaders interviewed
41 yrs
Deepest floor experience consulted
5
Structural parameters measured
The Problem

Quality control runs on eyes, after the metal has cooled

/ 01No data during the weld

Inspection happens after the metal has cooled

Today's quality control method: a trained inspector crawls into a ship's hull after the weld has cooled, camera in hand, and decides with their eyes. Zero data is captured between arc strike and inspection. By the time a defect is found, it's already inside finished metalwork.

/ 02The expertise is retiring

The expert workforce is disappearing

Reliable self-correction required welders with over a decade of intuition. That generation is retiring. Yards are filling roles with junior welders who haven't developed that judgment. The old system assumed expertise that no longer exists at scale.

The core problem

Zero data is captured between arc strike and inspection.

Arc strikeInspection
How It Works

Four steps. No change to the welder's process.

/ 01

Clip on

WarpSense attaches to any welding tool. No hardware swap, no process change.

/ 02

Measure in real time

Five structural parameters captured continuously while the weld is being made.

/ 03

Instant verdict

A pass-or-flag result the moment the joint is finished — before any inspector is needed.

/ 04

See what X-ray misses

Targets internal cracks between metal layers — invisible to the naked eye and frequently missed even by X-ray.

See It In Action

The dashboard that turns sensor data into a verdict

Our quality-scoring dashboard takes raw weld measurements and resolves them into a single pass-or-flag read — the same surface an inspector would use. Open it and explore a worked session.

WarpSense · Quality DashboardDemo
WarpSense quality dashboard: a roster of weld panels scored 0–100, colour-coded pass / needs-inspection, with fleet KPIs across the top.

A live, working interface — not a static screenshot.

Built full-stack from data pipeline to inspector alerts. Walk a scored weld session end to end.

Open the live dashboard →
Market Opportunity

A measurable wedge in a US$156B industry

TAM$4.7–7.8B

Global welding inspection services — 3–5% of the US$156B shipbuilding market, growing at 4.5% CAGR.

SAM$500M

Quality-related spend and failure cost in Southeast Asia — the fastest-growing NDT sub-region at 11% CAGR.

SOMS$17–22M

12–15 yards across Singapore, Malaysia, and Indonesia.

Addressable market narrowing to the SE-Asia beachhead
Stage 1

Beachhead

12–15 yards across Singapore, Malaysia & Indonesia · SGD 17–22M ARR.

Stage 2

Certification bodies

DNV and Bureau Veritas adoption triggers global portfolio roll-out.

Stage 3

Adjacent industries

Oil platforms, pressure vessels, and industrial fabrication.

Why No One Has Built This

A dataset moat that can only be built on the floor

/ 01

Hardware-agnostic by design

Equipment manufacturers can't build this without undermining their own products. We can.

/ 02

Camera systems are blind

Optical inspection misses sub-surface defects by design. WarpSense measures what cameras cannot see.

/ 03

A dataset moat

The proprietary dataset required — real weld measurements matched to real inspection outcomes — does not exist anywhere. It can only be built through live deployment. Every weld session adds to an advantage no late entrant can purchase.

We Went to the Yards

Field notes, not marketing copy

Three active production shipyards. Four named industry leaders across operations, inspection, project management, and commercial functions. Every conversation in person, on the floor.

It takes more than ten years for a welder to develop the judgment to know what is acceptable by eye alone.

Mr Tuck Wai · Head of Digitalisation, PaxOcean · 41 years in marine industry

The current quality control method: eyes and a camera for documentation.

Mr Lee · Certified Surveyor & Internal Inspector, PaxOcean

The highest-cost defect source is heat buildup between weld passes — invisible until the X-ray, and directly measurable in real time.

Mr Sam · Project Manager, LITA Ocean
The Team

Four founders, building on the floor

Ng Chen Meng

Business & AI Systems · NUS

Owns customer relationships and the AI verdict engine. Personally initiated and manages three active shipyard relationships.

Zhuzhen

Computer Science · NUS

Built the quality dashboard: raw sensor data to quality verdict. Owns the full stack from data pipeline to inspector alerts.

Yuan Yi Ning

Computer Science · NTU

Led every customer interview in person. The validation here carries real names and real quotes because she was physically in every conversation.

Sidhaarth Satish

Integrated Design & Engineering · NUS

Building the ruggedised sensor device from scratch — five simultaneous parameters, designed to survive arc flash, heat, and confined spaces.

The Plan

Raising SGD 150,000 to put sensors on live yards

A 12-month plan to move from concept to first paying contract. Forward-looking — every milestone below is a target, not a result.

  • Hardware36.67%
  • Software23.33%
  • Deployment20%
  • Operations & Legal13.33%
  • Contingency6.67%
Months 1–6
  • 3 sensors deployed across 3 live yards
  • First 1,000 real welds recorded
  • IP filing submitted
  • A*STAR research partnership initiated
Months 7–12
  • First paying contract live
  • SGD 1,500 / station / month (80 stations ≈ SGD 1.44M ARR per yard)
  • Certification-body conversation opened
  • Series A materials ready
Contact

Talk to us

Whether you operate a yard, work in inspection, or are looking at the space — we want to hear from you.