Flourish 1: Robot learns housework in 30 minutes
Flourish Robots is launching the Flourish 1, a humanoid household robot on wheels. You are supposed to be able to teach it tasks yourself via smartphone.
Tidying up shoes, loading the dishwasher, or watering plants: these are the kinds of household tasks the Flourish 1 is designed to take over. Instead of relying on a large library of pre-programmed routines, Flourish Robots focuses on individual training. The robot is meant to learn exactly how you want a task done in your own home.
How the training works
First, you define the individual steps in the app. For tidying up a shoe, this would involve driving to the front door, picking it up, and placing it in the closet. Flourish estimates this step takes about five minutes. You then guide the robot through the process several times using your smartphone, starting from different positions in the room.
According to the manufacturer, these demonstrations take about 30 minutes. About an hour later, the system should have processed the training and be able to perform the task independently. You do not need any programming skills for this.
In the process, the Flourish 1 gets to know the layout of your home. For example, it remembers where specific cabinets are located or where shoes should be placed. During later execution, no one controls it remotely. According to Flourish, the remote control is used exclusively for training purposes.
In addition to tidying up clothes, shoes, and toys, Flourish has already implemented clearing dishes. The robot can put cutlery and plates into the dishwasher, wipe tables, water plants, and fetch objects. It is also expected to be able to transfer laundry from the washing machine to the dryer.
You can start learned tasks via voice command or schedule them for specific times. Flourish cites the daily cleaning of a litter box at a set time as an example.
Source: Flourish Robotics
Learned routines can optionally be shared with other owners via a «Robot Store». Others can adopt these skills. Flourish has not yet explained how well this works in homes with different layouts.
Three cameras and LiDAR for orientation
The Flourish 1 moves on wheels and has a humanoid upper body with two arms. It is 1.09 meters tall and weighs about 20 kilograms. Its arms reach from the floor to a height of about 89 centimeters. The manufacturer specifies a maximum payload of about 1.5 kilograms.
Source: Flourish Robotics
For navigation, the robot uses a LiDAR sensor, which measures the distance to objects in its environment using a laser. Three cameras help it recognize and grasp objects. For this, the Flourish 1 uses image models that are trained based on your demonstrations.
Inside, it runs on a Raspberry Pi 5, which handles navigation, voice functions, and smaller AI models. More computationally intensive models run in the cloud.
The battery is expected to last about twelve hours. A full charge takes about nine hours. There is no automatic charging station for the first version.
Stairs and folding laundry are still too much for it
The Flourish 1 relies on individual trained routines. It does not scan a household independently, nor does it decide for itself which chores need to be done.
It cannot navigate stairs. With a payload of about 1.5 kilograms, heavier objects are also out of the question. It also cannot fold laundry yet. Flourish states that it is working on this capability.
Furthermore, its hands are not waterproof, so it cannot wash pots and pans. Flourish plans to include waterproof hands in a later version.
The first 50 units are scheduled to be available for pre-order starting September 29 for 3,555 US dollars. Delivery is planned for December 2026. Flourish has not yet provided any information regarding a wider market launch or availability in Germany, Austria, or Switzerland.
My interests are varied, I just like to enjoy life. Always on the lookout for news about darts, gaming, films and series.
From the latest iPhone to the return of 80s fashion. The editorial team will help you make sense of it all.
Show all
