Sunday Robotics

Overview

Sunday is a startup co-founded by Tony Zhao (author of the ACT paper, Stanford PhD dropout) and Cheng Chi (author of Diffusion Policy, Columbia/Stanford PhD student). They are gaining attention for their innovative approach of training without robot data, using only human motion data.

Item	Details
Headquarters	Mountain View, CA
Founded	2024
Co-Founders	Tony Zhao (CEO), Cheng Chi (CTO)
Funding	$35M (Benchmark, Conviction)
Robot	Memo

Founding Team

Tony Zhao (CEO)

Stanford PhD program (advised by Chelsea Finn, left in 2024)
UC Berkeley EECS B.S. (2021)
Lead author of ACT (Action Chunking with Transformers) paper
Developer of ALOHA / Mobile ALOHA robot systems
Previous: DeepMind, Tesla, GoogleX

Cheng Chi (CTO)

Started PhD at Columbia, moved to Stanford (advised by Shuran Song)
Lead author of Diffusion Policy paper
Developer of UMI (Universal Manipulation Interface)
Expert in robotics AI learning

Memo

Hardware Specs

Item	Specs
Type	Torso-style, wheeled base
Height	1.7m
Weight	77kg
Arms	2× multi-DoF
Gripper	Dual
Height Adjustment	Floor to 2.1m

Capable Tasks

Dish organization and dishwasher loading
Handling wine glasses (delicate objects)
Espresso brewing
Laundry folding

Zero Robot Data Approach

Skill Capture Gloves

Using $200 gloves instead of robot teleoperation:

Item	Teleop Rig	Sunday Gloves
Cost	~$20,000	~$200
Location	Lab	Real homes
Collectors	Experts	”Memory Developers”
Data	Robot motions	Human motions

Data Pipeline

[Human wearing gloves] → [Performing chores at home] → [Motion data collection]
                                                              │
                                                  ┌───────────▼───────────┐
                                                  │   Skill Transform     │
                                                  │ (Human→Robot motion)  │
                                                  └───────────┬───────────┘
                                                              │
                                                  ┌───────────▼───────────┐
                                                  │       ACT-1 Model     │
                                                  └───────────────────────┘

Data Scale

Item	Value
Gloves Distributed	2,000+
Data Collection Homes	500+
Household Task Episodes	10M+

ACT-1 Model

Features

Item	Details
Core Innovation	Zero Robot Data
Training Data	Human motions only
Combined Features	Manipulation + Navigation
Generalization	Zero-shot to new environments

Skill Transform

Software for converting human motion → robot motion:

Claims 90% success rate
Human dexterity → Robot gripper mapping
Includes force/contact data

Deployment Plans

Timeline	Plan
2025.11	Memo launch, beta applications begin
Late 2026	”Founding Family” beta (~50 households)
Current Production Cost	~$20,000 (handmade)
Target Retail Price	<$10,000 (at scale)

Significance

If Sunday’s approach succeeds:

Problem	Solution
Data Bottleneck	Large-scale collection via affordable gloves
Environment Diversity	Collection in real homes
Cost	Home robots under $10,000
Generalization	Zero-shot adaptation to new environments