Insulation that looks sound from the outside tells you almost nothing about the steel underneath. Moisture migrates inside the system, so corrosion often appears a long way from wherever the water entered. That is the bind every CUI program works inside: the damage is hidden and localized, which is precisely the combination that sample inspections handle worst.
Scaffolding, insulation removal, disposal and reinstatement carry most of the cost of a traditional inspection, so programs inspect what they can afford to reach and infer the rest. Risk-based inspection still governs the decision, but the evidence feeding it is minimal.
Robotics changes the economics. Clamping robots and balancing crawlers cover hundreds of feet of pipe per day; a robotic arm works from grade and reaches vessels as well as piping; a UAV goes where nothing can attach. What rarely gets addressed is what each robot doesn’t accomplish.
Petter Wehlin, product manager at ARIX® Technologies, addresses four robot applications using eight criteria in the July/August 2026 Inspectioneering Journal: pipe orientation, inspection coverage, NDT payload, obstacle traversal, clearance, productivity, operational complexity and inspectable assets.
Wrap the pipe and drive along it at any orientation, rotating a full 360 degrees and carrying PEC and RTR at the same time. That is the only configuration in the set that screens and quantifies in a single pass. The cost is specificity: each robot is built for a diameter range, and flanges, valves and elbows still need a manual inspection.
Ride the top of horizontal runs on tracks that traverse across damaged or uneven cladding. They cannot rotate, so coverage is fixed to preset azimuths such as the 6 o'clock position where water collects, and horizontal pipe is the whole of their range.
Reach from grade into any orientation, piping and vessels alike, without touching the asset. The tradeoff is throughput and setup: larger machines, higher power draw, careful base placement, and lower productivity per foot than crawlers and clamping robots.
Reach anything with approach clearance, including assets nothing can clamp to, and return PEC spot readings rather than continuous scans. RTR is too heavy to fly, weather and airspace rules apply. A UAV cannot be shut down during a gas release, which matters in a congested rack.
Full coverage means entire circuits get screened systematically, and findings are located with higher confidence.
Complex geometry and congested racks still leave gaps, which is why the paper argues for a hybrid program by design: robotic scanning across the accessible majority, other methods against the remainder. Treating any single platform as the answer is how coverage claims get overstated.
Curious what this looks like on your unit? A field walk puts ARIX on site to scope what robotic inspection reaches, and what it does not. Book a Field Walk.