Getting ahead of Apple’s plan to enter the robotic device market, Samsung's robot branded “Ballie” will go on sale in the US and South Korea this summer, the company announced today. What's more, through a partnership with Google Cloud, the diminutive robot will ship with a Gemini AI model.

According to Samsung this morning, “Ballie will use Gemini’s multimodal capabilities along with proprietary Samsung language models to process and understand a variety of inputs, including audio and voice, visual data from its camera and sensor data from its environment. This will allow it to adapt its behaviors and responses in real-time. For instance, if you’re not sure what to wear to work in the morning, you can turn to Ballie for advice. Users could ask, “Hey Ballie, how do I look?” and Ballie can respond with styling recommendations, such as trying new accessories or adding a colorful shirt.

With enhanced reasoning enabled by Gemini, Ballie will be able to help users manage not only their home, but also their health and wellbeing. For example, a user could tell Ballie, “I feel tired today.” Ballie could then use Gemini’s grounding in Google Search to give tailored advice on how to improve their energy levels, including sharing recommendations from trusted sources on how to get more exercise, optimize their sleeping environment or monitor their sleep patterns.

This partnership announcement builds on the successful integration last year of Gemini on Google Cloud into Samsung’s Galaxy S24 smartphone series. With Ballie, Samsung and Google Cloud are exploring how to bring their generative AI collaboration to a wider range of devices.

“With Gemini on Google Cloud, Samsung is demonstrating how to deploy generative AI at scale, integrating it directly into the heart of their popular products,” said Thomas Kurian, CEO, Google Cloud. “This expanded partnership with Samsung underscores our commitment to delivering reliable, enterprise-grade AI that drives lasting value for our customers – and their customers.”

Samsung has yet to announce pricing for Ballie, though with the uncertainty around the Trump administration's recently announced tariffs, you can't blame Samsung, can you? The company first demoed Ballie in 2020. The robot has gone through a few iterations since the then.

Earlier this month, the World Intellectual Property Organization published  Samsung’s patent WO2025071029 that directly relates to their robot branded “Ballie.”

According to Samsung’s patent, an electronic device according to an embodiment comprises: a projection device for projecting an image; a driving device for adjusting a projection direction of the projection device; a memory for storing at least one instruction; and one or more processors connected to the projection device, the driving device, and the memory so as to control the electronic device, wherein the one or more processors control the projection device to project an image in a first direction, and when a predetermined event occurs, control the driving device such that an image is projected in a direction different from the first direction.

The electronic device may further include a camera, and the one or more processors may control the camera to photograph a space around the electronic device, and may obtain a plurality of projection areas on which the projection device projects an image based on the captured surrounding space.

The one or more processors may obtain a first projection area for a space located above the electronic device and a second projection area for a space located on the floor.

The one or more processors may have a predetermined size of a bottom surface, and may obtain a second projection area in a space having a predetermined size from among a bottom surface located between the user and the electronic device.

The one or more processors may identify a wall surface on the captured surrounding space, and may obtain one or more first projection regions by identifying a region having a preset color when there is no obstacle among the identified wall surfaces.

The one or more processors may control the driving device to change the projection direction of the projection device from the first projection area to the second projection area when the need to interact with the image is identified.

The one or more processors may identify a user selection area among projected images by using an image captured through the camera during projection of an image to the second projection area, and may perform an event corresponding to the identified user selection area.

The one or more processors may control the driving apparatus and the projection apparatus to select the first projection area or the second projection area and project an image to the selected projection area.

The one or more processors may control the driving device and the projection device such that a user interface window for receiving a user setting is projected to a second projection area when at least one of the setting of the electronic device or the reproduction environment of the image is required.

The one or more processors may identify a direction indicated by a user by using an image captured by the camera, and obtain a space corresponding to the identified direction as a projection region.

The electronic device may further include a microphone, and the one or more processors may recognize a user voice input through the microphone and control the driving device to change the projection area based on a user voice.

The electronic device may further include a camera, and the one or more processors may control the driving device to analyze the posture of the user based on the captured image when the user is photographed through the camera, and change the projection area based on the posture of the user.

The control method may further include, when an image captured by the user is input, identifying at least one of a posture of the user and a gaze of the user by analyzing the input image, and the changing may include changing a projection direction to project an image to a second projection area corresponding to the identified user posture and the user gaze among the plurality of projection areas.