Commercial pool skimmer robots run between $800 and $1,800 and are mostly closed black boxes — proprietary chargers, no parts catalog, batteries that fail at month 18 with no path to swap.
Commercial pool skimmer robots run between $800 and $1,800 and are mostly closed black boxes — proprietary chargers, no parts catalog, batteries that fail at month 18 with no path to swap. The path that DIY builders keep landing on is two outrunner BLDC thrusters from the e-foil/e-surfboard parts ecosystem, a foam or HDPE hull, and an ESP32 + GPS for surface navigation. Total bill of materials lands between $250 and $500 depending on whether you reuse a Li-ion pack you already have.
The hard part is not propulsion — marine BLDCs at this scale are commodity now.
The hard part is not propulsion — marine BLDCs at this scale are commodity now. The hard part is path planning that doesn't require a real LIDAR. The trick is that pools have known finite geometry: you teach the robot the shape once (manual joystick lap with GPS-RTK or a phone NTRIP correction) and after that it just zigzags inside the polygon and uses the downward camera plus an IMU to debias drift. The mesh net under the hull catches everything between 2mm and 50mm — leaves, dead bugs, hair clumps — and a wall-side dock with a USB-C charger is the place where you empty it once a week.
If you want to skip the GPS-RTK route, the alternative is four magnetic markers on the pool corners: the robot triangulates against them with a coil-loop sensor and stays inside the box.
If you want to skip the GPS-RTK route, the alternative is four magnetic markers on the pool corners: the robot triangulates against them with a coil-loop sensor and stays inside the box. We've seen builds work with both approaches; magnetic is cheaper, GPS is more general (works on uncovered pools where the rim isn't a clean reference).