Rust-Removal Robots: It's Not About Water Pressure, It's About Positioning
Community Discussion · Humanoid Robots

Rust-Removal Robots: It's Not About Water Pressure, It's About Positioning

Mo MoMo Mo2d ago2026/11/17 85 views

Let me say something. I've been digging through a lot of material on ship rust-removal robots recently — no chance to actually get on a dry dock, but reasoning backward from papers and public delivery cases, the more I look the more I feel the marketing angle has grabbed the wrong focus. The promotion is all about ultra-high-pressure water jet pressure, cleaning width, magnetic adhesion force. These are of course critical, but they don't determine whether the machine can hold its ground in a dry dock. What determines that is coverage rate.

The evidence is in all those path-planning papers. In theses on this kind of wall-climbing robot, the heaviest part is wall-walking control — fused localization, point-to-point path planning, complete coverage path planning — while the nozzle barely gets a few pages. The hull's outer plating is a doubly curved surface; once the magnetic crawler track slips, the wheel odometry starts drifting. This matches what I've felt these past two weeks messing with IMU plus odometry — fine over short distances, but once you go out a dozen-plus meters, the pose error is enough to make you re-clean the spots you already did.

Re-cleaning isn't just a waste of time. Ultra-high-pressure water hitting steel plate has a cost — every extra pass is another risk of surface damage, and another batch of wastewater and paint residue to deal with.

Recovery is the real threshold. AKA Robotics recently delivered a drillship maintenance system in the Gulf of Guinea off West Africa — magnetic crawling with ultra-high-pressure water jet, paired with wastewater and paint-residue recovery equipment. It looks like an environmental bonus, but it's actually an entry requirement. If paint residue flows into the sea, you won't pass the environmental gate, and no matter how good the machine is, it won't get on site.

So my judgment is one sentence: this industry competes on how you know where you've been, and nozzle pressure comes after that. If localization drifts, coverage has gaps; if coverage has gaps, you need manual touch-up — and once manual touch-up comes in, the labor you saved goes right back. China Ship News visited Shihe a couple weeks ago, which shows someone really is pushing this line seriously, but hopefully the direction they push isn't just the pressure gauge.

2 replies

?
Ctrl + Enter to reply
Dexterous Hand Controller

Drifting a dozen meters means you have to redo the subtraction all over again—this matches exactly what I felt tuning IMU + odometry. When coverage misses, manually patching it is the real cost.

Demo Still Far
Reply to Dexterous Hand Controller

You're underestimating the cost of touch-up painting. Drift and re-stripping isn't just labor hours—running ultra-high-pressure water over the steel plate again doubles both plate damage and paint sludge disposal volume. The recovery unit is the first to buckle.