ONE SYSTEM.
TOTAL
VISIBILITY.

AX-1 — the first autonomous inspection crawler purpose-built for insulated cast iron pipelines in oil & gas and industrial environments.

AccuBotix crawler inspection angle 2
AccuBotix crawler inspection angle 3
AccuBotix crawler overview
AccuBotix monitoring ecosystem
AccuBotix inspection report output
The Problem We're Solving
Pipelines fail silently. doesn't let them.

THE OLD WAY

  • Manual inspection teams access one segment at a time
  • Insulation removed, inspected, re-installed — weeks of downtime
  • No detection until corrosion is already severe
  • Human error, inconsistent data, no historical record
  • Reactive maintenance — fix what already broke
  • High cost, high risk, low coverage

THE WAY

  • Autonomous crawler traverses live pipelines continuously
  • Inspects through insulation — zero removal required
  • Detects defects at sub-millimetre precision, in real time
  • Every scan geotagged, timestamped, stored in digital twin
  • Predictive failure modelling — act before damage occurs
  • Fraction of the cost. Zero unplanned shutdowns.
Why
Three sensors. One truth.

Most inspection tools see one thing. fuses visual, thermal, and acoustic data through a Bayesian AI engine — cross-validating every finding to eliminate false positives and deliver a ranked, actionable repair list.

👁

Visual AI

360° 4K ring camera with CNN-powered crack and corrosion detection. Catches surface anomalies invisible to the human eye at speed.

🌡

Thermal Sensing

LWIR microbolometer detects leaks and blockages through insulation by mapping temperature differentials as small as 1°C — no stripping required.

📡

Acoustic Emission

Dual-mode AE + ultrasonic sensor pinpoints active cracks via TDOA localisation and measures wall thickness to ±0.1mm, catching hidden corrosion early.

Edge AI

NVIDIA Jetson Orin NX processes all three sensor streams onboard at >5 FPS per module. No cloud dependency. No latency. Results in real time.

🗄

Digital Asset Twin

Every defect tagged to a Segment ID with precise distance markers. Full material, age, and repair history per pipe section — enabling predictive RUL modelling.

📊

Auto-Reports + CMMS

PDF inspection reports auto-generated post-run with ranked repair priorities. Direct CMMS/ERP work order triggers on threshold breach. Zero manual admin.

"The question isn't if your pipeline will corrode. It's whether you'll know about it before it fails."

AccuBotix— Built for Oil & Gas, Petrochemical & Industrial Plant Operators

Use Cases
Built for the toughest pipelines on earth.
01
Upstream Oil & Gas

Wellhead flow lines, gathering networks, and offshore topsides — inspected at full diameter without production shutdown.

02
Refineries & Petrochemical

Process piping running hot fluids under high pressure — continuous wall-loss monitoring under lagging insulation.

03
LNG & Gas Transmission

Cryogenic and high-pressure lines where condensate blockages and external corrosion are constant threats.

04
Industrial Plants

Steam distribution, cooling water, and utility networks in power plants, chemical facilities, and manufacturing.

AX-1

AX-1 Crawler

Autonomous Pipeline Inspection Platform

The AX-1 is a ruggedised 4WD autonomous crawler designed to traverse DN100 cast iron pipelines under live operating conditions. Spring-loaded wheel assemblies provide constant gripping force through aluminium cladding on inclines up to 30° and full vertical risers. The articulated chassis hinge enables navigation through 90° and 45° elbows without requiring pipeline shutdown or insulation removal.

Target Pipe
DN100 Cast Iron (4")
Drive System
4WD Spring-loaded wheels
Max Incline
30° continuous
Vertical Capability
1 m vertical riser
Bend Traversal
90° + 45° elbows
Battery Life
~3.5 hours continuous
Compute
NVIDIA Jetson Orin NX 8GB
Storage
Dual 512 GB industrial microSD (mirrored)
Odometry
Quad encoders + MPU-9250 IMU
Position Accuracy
Sub-centimetre (RFID + dead reckoning)
RFID Resets
At every pipe joint
Communication
Wi-Fi offload post-run
No Pipeline Shutdown Inspects Through Insulation Zero Insulation Removal
SEN

Trimodal Sensor Head

Visual · Thermal · Acoustic — Simultaneous

The sensor head mounts directly to the AX-1 chassis and captures all three modalities simultaneously — visual, thermal, and acoustic — feeding independent AI modules running in parallel on the Jetson Orin NX. Every sensor reading is stamped with Segment ID and precise pipe distance for full spatial traceability.

Visual
4K 360° ring camera array
AI Model — Visual
CNN crack / corrosion detection
Thermal Sensor
LWIR uncooled microbolometer
Thermal Spectrum
8–14 μm
Thermal Sensitivity
≥1°C differential
AI Model — Thermal
Statistical anomaly engine
Acoustic Mode 1
40 kHz AE — crack / leak TDOA
Acoustic Mode 2
Pulse-echo UT thickness gauging
Wall Thickness Resolution
±0.1 mm
AI Model — Acoustic
FFT + LSTM classifier
Fusion Engine
Bayesian multi-modal cross-validation
Inference Speed
>5 FPS per modality
SFT

Cloud Platform

Dashboard · Reports · Digital Twin · CMMS Integration

The platform handles everything from raw sensor data to boardroom-ready inspection reports. Upload completes in under 30 minutes. Every pipe segment builds a living digital twin — tracking material grade, operating conditions, defect history, and remaining-useful-life estimates across all inspection runs.

Cloud Storage
AWS S3 / Azure Blob
Offload Time
<30 min for 3.5 hr session
Reports
Auto PDF — defect table + images + repair priority
Integrations
CMMS / ERP work order triggers
Digital Twin
Per-segment asset database
Predictive
RUL modelling + failure probability forecasting
Middleware
ROS 2 Humble
Licensing
SaaS subscription or on-premise deployment

Open-Source Software Stack

Tool Purpose
ROS 2 HumbleRobot middleware, sensor topic bus
OpenCV 4.xVisual crack and corrosion detection
PyTorch / TensorRTCNN vision model + Jetson inference
Mosquitto MQTTLightweight sensor data broker
Node-REDCloud pipeline and dashboard wiring
Gazebo HarmonicDigital twin physics simulation
Request Pricing & Demo →

READY TO SEE THE AX-1 IN ACTION?

Schedule a Demo