SwitchBot Debuts Embodied AI Lineup with Emotional, Visual, and Athletic Intelligence

• SwitchBot introduced the AI Pet, AI Hub, and AI Tennis Robot as part of its embodied AI showcase at IFA 2025.
• Each device brings real-time perception and personality into daily routines, blending emotion, vision, and movement.

At IFA 2025, SwitchBot unveiled three new products built around embodied AI, including a Tennis Robot, an AI Pet, and an AI Hub. Each device is designed to sense, respond, and interact with people and environments in real time, combining physical robotics with on-device or cloud-based intelligence.

The Acemate Tennis Robot is a mobile training system that uses dual 4K binocular cameras and AI to track serves, returns, and rallies with centimeter-level accuracy. It also functions as a tennis coach, capturing spin, speed, net clearance, and placement data with feedback delivered in-session through the Acemate app. It supports 20 programmable target zones, multiple serve modes, has an 80-ball capacity, and up to three hours of play on a single charge.

The SwitchBot AI Pet is a soft-bodied household robot designed for emotional companionship. It features on-device large language model AI and cloud-based visual language model AI to recognize emotions and respond in context. It expresses emotions such as happiness, loneliness, jealousy, or hunger, and learns from daily interactions. The AI Pet remembers people, routines, and spaces, keeping a log of meaningful moments to create a more personal connection over time.

Finally, the SwitchBot AI Hub is the first smart home edge hub with a Vision Language Model AI. Paired with devices like the Pan/Tilt Cam Plus 2K/3K or Smart Video Doorbell, it can visually interpret events and generate text summaries that act as triggers for automation. It supports text-based searches like “Show me when I left my phone” and delivers daily summaries through the SwitchBot app. The hub connects with over 100 devices and features a built-in 6T AI chip, making it an entirely local, vision-aware command center for the connected home.

🌀 Tom’s Take:

The leap from voice to vision is reshaping what AI can do, making devices not just reactive, but situationally aware. VLMs turn footage into function.

Source: PR Newswire / Switchbot