A swimming pool may seem, at first glance, one of the simplest settings for a cleaning robot: a relatively delimited surface that you only have to navigate. But underwater things get complicated. There are depth changes, steps, drains, walls, obstacles and debris ranging from tiny grains of sand to leaves. Besides, The robot not only has to move along the bottom: it must also know when and how to climb walls or get close to the waterline.
That is the territory that Roborock has chosen to expand its commitment to domestic robotics. The company, known for its robot vacuum cleaners, presented the RockAqua P1 at IFA 2026, its first pool cleaning robot and also the first member of a new family of devices intended for automated pool maintenance. The proposal is based on a simple idea: that Cleaning a pool does not consist of mechanically crossing its surface, but rather the robot being able to interpret the environment and adapt its behavior to it.
The RockAqua P1 is designed to work in different pool configurations and cover not only the bottom, but also the walls, waterline and shallow platforms, with a depth of up to 15 centimeters. To move through these spaces, it uses an adaptive cleaning trajectory and a vision-based obstacle avoidance system, capable of identifying elements such as steps and drain covers.
It is not a minor detail. A robot that simply follows a predetermined path may find that a real pool has little to do with a perfect grid. Obstacles force you to change direction and the different shapes of the glass can leave areas uncleaned. Intelligent navigation aims precisely to reduce these interruptions and ensure that the robot dedicates its time to cleaning. and not to repeat trajectories or get stuck.
It also changes the nature of the dirt it has to deal with. Very fine particles, sand, small debris and larger leaves can coexist on the floor of a pool. To cope with that fan, The RockAqua P1 combines a water circulation capacity of more than 25,700 liters per hour, with a four-liter waste basket and a double filtration system, with 180 and 70 micron filters.
Another important detail is that it does not maintain the same suction intensity throughout the entire journey. When you encounter large debris, you can temporarily reduce the suction to prevent the movement of the water from displacing them before trapping them and increase it again when it approaches the dirt. It is a strategy that tries to adapt the robot’s strength to what is in front of it, instead of always using maximum power.
Added to this capacity is ClearVision Smart Patrol Cleaning, a system that allows prioritizing the areas that have the greatest accumulation of dirt. The goal is to avoid two fairly common problems in automated cleaning: some areas receiving too many passes while others are left practically untouched. Autonomy is another challenge when the space to be cleaned can be considerable. The RockAqua P1 offers up to 3.5 hours of effective operation for cleaning the floor, while its motor management adapts energy consumption to the conditions of the pool and the amount of waste it finds.
But perhaps the most curious part of the operation appears when the work is finished. Instead of leaving the owner with the task of locating the robot somewhere on the bottom, the RockAqua P1 can automatically return to the surface when it completes its cleaning cycle or when the battery reaches a critical level. There it remains waiting to be picked up. The recovery system also takes into account something as unspectacular as the weight of the water. When removing a robot from a pool, much of the effort comes from having to lift a device still loaded with water. The RockAqua P1 incorporates a quick emptying system that allows it to be evacuated when removed, making it easier to transport to the charging base.
Control is completed through an application from which it is possible to check the status of the device and configure cleaning routines. Thus, the objective is not only to automate the movement of a device underwater, but to convert pool maintenance into a task that requires less and less intervention. With the RockAqua P1, Roborock thus enters a category in which navigation intelligence takes on a different meaning than that of domestic robot vacuum cleaners. On the floor of a house, an obstacle can be a chair or a wall; In a pool, it can be a step that changes the geometry of the path, a drain cover or a transition between different depths. The pool thus becomes a new testing ground for cleaning robotics: a three-dimensional space in which the robot has to orient itself, detect obstacles, manage different types of dirt and, finally, even know how to return to the surface so that its owner can recover it.