Sunday is developing a general-purpose home robot, Memo, with the ambitious goal of placing over a billion units in homes to eliminate domestic toil.
The company's core strategy is a full-stack approach, combining proprietary hardware with an AI model trained on a massive, real-world dataset collected via custom data-gloves.
Sunday's key innovation is a scalable data collection method that favors real-world imitation learning over simulation, having already gathered nearly 10 million long-horizon trajectories.
The company is targeting a 2026 beta program and a potential commercial launch in 2027-2028, aiming for a sub-$10,000 price point at scale.
10 quotes
Concerns Raised
Scaling the full-stack robotics system is acknowledged to be extremely difficult, with many unforeseen mechanical, data quality, and reliability issues.
Maintaining high-quality, precise data collection at scale from hundreds of non-expert users presents a significant operational challenge.
The bar for a consumer product is incredibly high, requiring extreme safety, high capability, and a low cost point to achieve mass adoption.
Opportunities Identified
The total addressable market for a general-purpose home robot that can eliminate domestic chores is enormous.
Their novel, real-world data collection methodology creates a powerful and defensible moat against competitors.
Robotics performance is expected to improve dramatically with data and model scale, following a similar trajectory to large language models.