Sunday Unveils Memo: A New Household Robot Designed for Real-World Tasks

On November 19, 2025, the Mountain View-based startup Sunday introduced Memo, a personal robot aimed at assisting families with household chores. Developed by Stanford-trained roboticists Tony Zhao and Cheng Chi, Memo distinguishes itself by learning from actual domestic behavior rather than relying on industrial simulations or controlled lab environments. This innovative approach addresses a significant gap in home robotics, as highlighted by the company.

Interesting Engineering

The core of Memo's development lies in its extensive data collection from over 500 households, utilizing a patented Skill Capture Glove. This wearable device recorded how individuals move, clean, sort, and organize objects, resulting in nearly 10 million episodes of real household routines. This dataset is touted as unmatched in diversity and scale, enabling Memo to perform complex, multi-step tasks that require context and decision-making. Tasks such as clearing tables, running dishwashers, folding laundry, sorting shoes, and brewing espresso are now within Memo's capabilities, which have traditionally posed challenges for robots trained solely in structured environments.

Interesting Engineering

Memo's design deviates from the trend of humanoid robots, opting instead for a wheeled base that enhances mobility and stability. Its central column can adjust its height, allowing it to perform tasks at various levels while maintaining balance, even during power loss. The robot features a glossy white body, a friendly cartoon-like face, and interchangeable colorful baseball caps, embodying a retro-futuristic aesthetic reminiscent of characters like Baymax from "Big Hero 6." This design aims to make Memo a soft and approachable addition to home environments rather than an industrial machine.

Interesting Engineering

Sunday plans to launch an early access program for Memo, named the Founding Family Beta, starting on the same day as the announcement. Fifty households will be selected to receive a numbered unit and direct support, allowing Sunday to refine Memo's capabilities before a broader release. The startup has expanded to a team of 25 engineers and researchers with backgrounds from notable institutions such as Stanford, Tesla, DeepMind, Waymo, Meta, and Neuralink.

Interesting Engineering

Tony Zhao emphasized the importance of real-world data in developing effective home robots, stating, “The problem has always been data. Most home robots start as adaptations of industrial machines, and those trained in labs rarely succeed in unpredictable, real-world environments.” Eric Vishria, General Partner at Benchmark, echoed this sentiment, noting that the promise of AI robotics lies in their ability to function in messy, real-world situations.

Interesting Engineering

While Memo's innovative approach has garnered attention, some experts remain skeptical about the practical implementation of household robots. Critics argue that despite advancements in data collection, the unpredictable nature of home environments may still pose significant challenges for robotic assistance.

Interesting Engineering

As Sunday prepares for the early access program in 2026, the company aims to gather feedback from initial users to further enhance Memo's functionality and reliability, paving the way for a wider market introduction.

Interesting Engineering