Want a robot vacuum but don’t have the budget or don’t want to buy one from commercial brands? A recent open-source project called OOMWOO lets you fulfill your dream of building your own robot vacuum from scratch. The OOMWOO robot vacuum is fully open source from hardware to software, with a focus on requiring no cloud services whatsoever. It uses a Raspberry Pi as its core, paired with a 2D LiDAR sensor for home mapping and autonomous navigation, and comes with native Home Assistant integration built in, allowing users to fully control the robot vacuum locally without handing over all their data and remote control to a remote server.
OOMWOO builds their own robot vacuum with Raspberry Pi and 2D LiDAR
According to the official website and GitHub project documentation, OOMWOO’s core components include Raspberry Pi, ROS2 navigation system, 2D LiDAR mapping, and Home Assistant integration, with the chassis featuring a modular structure primarily based on 3D printing. The electronics also utilize ESP32 and Arduino, and the overall project is released under the Apache-2.0 open-source license; hardware design files are also released under an open hardware license.
The key to keeping costs down while maintaining functionality lies in the fact that this machine abandons the top-tier sensing solutions commonly found in commercial robot vacuums, instead using a fairly cheap 2D LiDAR system to achieve the same mapping and obstacle avoidance. The development team noted in the design documents that the goal is to create a home appliance with “product-grade quality” rather than a disposable assembly toy, while keeping it easy to assemble, with step-by-step instructions for beginners, reproducible via 3D printing, and with parts that users can source themselves.
For those who want to try it early, the project also offers a path that doesn’t require waiting for hardware: you can connect the off-the-shelf Proscenic M6 Pro robot vacuum to ROS2 and use it as a temporary stand-in before the official version is ready, running the same navigation and mapping stack on real hardware. The development environment also supports testing SLAM, Nav2, and remote control in the Gazebo simulator first, so even people without a robot can get hands-on experience with the whole system.
Current progress: still in early development
However, it’s worth noting that OOMWOO is still in the early development stage. The official website labels it as “Early development,” and the first Bill of Materials (BoM) is expected to be released around mid-July 2026 (which hasn’t appeared yet). A formal, complete cost estimate won’t be available until the end of August that year, with step-by-step assembly tutorials expected to be published in the fall. From the reference design diagrams released on the official website, the bottom of the unit is equipped with drive wheels, brushes, omni-wheels, and cliff sensors; lifting the top cover reveals the LiDAR, battery, cooling fan, and an accessible I/O mainboard. The structural design leaves considerable room for disassembly and modification.
The software side is progressing faster, with most development work already complete. This includes the development environment, ROS2 simulation, connection tutorials for temporary substitute hardware, and a rough bill of materials, all of which are already live. Many remaining items still await community contributions, such as mapping 3D printing files to off-the-shelf parts, Raspberry Pi software packages, I/O circuit boards and firmware, and complete assembly and debugging tutorials. The entire project is developed in parallel using an “RFC module” approach, where volunteers can claim modules they’re interested in (simulation, mapping, navigation, charging dock, floor cleaning, etc.), develop in their own repos, and then link back to the main project via PRs.
OOMWOO project URL / Official website
Conclusion
In terms of positioning, OOMWOO is not yet a mature product you can buy and put to work at home right away; it is more like an open-source platform that invites everyone to get hands-on. Its significance is that it shows the market another possibility: robot vacuums don’t have to be closed black boxes that are either bought outright or subscribed to—they can also be DIY-assembled toys, though it does cost considerable effort, time, and the money for the parts and a 3D printer.
Source: KOCPC Chinese