Flourish Unveils Budget Home Robot Taught via Smartphone

Recap

Source: 로봇신문, report of Oct. 5, 2026

According to the report, Flourish, a San Francisco-based robotics startup, has unveiled the Flourish One, a home robot designed to learn household tasks through smartphone demonstrations. The device features a wheeled chassis with two arms, each capable of lifting up to 1.5 kg, and is equipped with a small LiDAR sensor for obstacle detection and cameras on the gripper wrists for environmental awareness. The company states that the robot can handle simple chores such as organizing items, taking out trash, and wiping tables.

To keep costs down, Flourish uses a Raspberry Pi as the local computing platform instead of a high-performance GPU, relying on cloud-based GPU resources for AI training and high-level inference. The company says that installing an NVIDIA GPU in the robot would significantly increase the price. The initial production run consists of 50 units, priced at $3,555, with the goal of shipping the first units before Christmas.

The core interaction model involves a smartphone app that uses the device's IMU to capture user movements. Founder Antoine Marcel explains that users can teach the robot by performing tasks with their hands while holding the phone; the robot then mimics these motions. Marcel notes that this approach allows for task-specific fine-tuning of the AI model for each household, which he argues is more practical than applying a single universal AI model to every home.

Context

The home robotics market has largely focused on high-end, general-purpose humanoid robots that command premium prices due to complex hardware and energy requirements. Flourish’s approach diverges from this by prioritizing affordability and task-specific utility over general intelligence. By offloading heavy computational loads to the cloud and using low-cost local hardware, the company aims to make robotic assistance accessible to a broader consumer base. This strategy contrasts with competitors who often embed powerful local processors to ensure offline capability, a trade-off that typically drives up the unit cost.

Robot's take

Flourish One represents a pragmatic shift in home robotics, betting that specialized, low-cost units can find a niche where expensive generalists fail. However, the reliance on cloud connectivity for core AI functions raises questions about latency and privacy, especially in a home environment. The 1.5 kg payload limit and wheeled design also restrict its utility compared to bipedal systems. It remains to be seen whether the smartphone-based teaching interface will be intuitive enough for non-technical users to adopt widely.