Figure Gen 3 Robot Launch: Handles Laundry, Dishes, and All Housework — Netizens Say Non-Humanoids Have No Future
Figure 03 — The Third Generation Humanoid Robot
The Figure humanoid robot family has a new member — the third-generation Figure 03.
According to the official statement, this model is purpose-built to integrate with the Helix AI system, fit seamlessly into home scenarios, and meet the requirements for large-scale global deployment.
---
Demonstrated Home Chores & Autonomy
Figure 03 shows clear logical reasoning in home chores, such as:
- Tidying spaces — placing toys into a basket, closing and placing a laptop neatly.
- Video link
- Cleaning tableware — removing food scraps nearby.
- Washing dishes — rinsing scraps under running water, loading plates into a dishwasher.
- Cup storage — turning cups upside-down before storing.
- Laundry handling — adding detergent, setting washing machine, starting cycle.
- Video link
- Folding clothes neatly and putting them away.
---
Autonomy Debate
Netizens debated which tasks were remote-controlled vs. fully autonomous:
> “If everything is truly self-directed (no remote operation, no pre-scripted demo), it’s a major milestone.”
> “The way it pats the pillow after placing it hints at remote control.”
CEO Brett Adcock’s direct response:
> “Remote control level: zero.”
---
Rapid Growth and Funding Highlights
Backed early by OpenAI, Microsoft, NVIDIA, and others, Figure has achieved:
- Funding exceeding USD 1 billion
- Post-money valuation: USD 39 billion — highest disclosed in the embodied intelligence sector.
---
Hardware & Software Redesign
Since ending its partnership with OpenAI in Feb 2024, Figure has accelerated progress.
Core goal:
> To use the Helix system for true autonomous decision-making.
Figure 03 upgrades include:
- Fully redesigned sensor suite and hand system
- New-generation vision system for frequent vision–motion control scenarios.
Vision system improvements:
- Double frame rate
- 75% latency reduction
- 60% wider field of view per camera
- Extended depth-of-field for denser perception data streams
---
Hand Design & Embedded Palm Cameras
Palm cameras in each hand:
- Wide FOV, low latency
- Maintain visual perception when main camera is obstructed
---
Finger Flexibility & Tactile Sensing
- Softer, adaptive fingertips for greater contact area and stability
- Proprietary first-gen tactile sensors — ultra-durable, high-fidelity
- Pressure detection down to 3 grams (≈ weight of a paperclip)
---
High-Speed Data Offload & Machine Learning
- 10 Gbps millimeter-wave data offload for terabyte-scale uploads across robot clusters
- End-to-end “pixels-to-action” learning capabilities
---
Home-Oriented Safety Optimizations
- Multi-density foam pinch-point protection
- Soft, removable, washable exteriors
- Optional cut-resistant, durable coverings
Form factor:
- 9% lighter than Figure 02
- More compact for home agility
---
Battery Protection & Audio Upgrades
- Multi-layer battery safety design meets UN38.3 standards
- Audio improvements:
- Double-sized speakers, 4× power output
- Repositioned microphones for clarity
- Video link
---
Wireless Charging & Data Offload
- Wireless induction charging via foot coils — 2 kW fast charge
- Autonomous docking in home scenarios
---
Design for High-Volume Manufacturing
- Simplified component design & assembly
- Transition from CNC machining to die casting, injection molding, stamping
- Significant cost reduction per unit with volume scaling
- Vertical integration of actuators, batteries, sensors, structures, electronics
Partner supply scalability:
- Ability to produce “tens of thousands to millions” of parts
BotQ Manufacturing Facility:
- Capacity: 12,000 robots/year initially
- Target: 100,000 units/year within 4 years
---
Commercial Advantages
- Double actuator speed, improved torque density
- Upgraded Helix-adaptive hands & sensors
- Inductive charging + high-speed wireless data offload
- Customizable exteriors for branding & durability
- New side display for quick identification in clusters
---
Industry Viewpoints
Philip Johnston (Starcloud CEO):
> “Other forms of robots will be defeated, aside from humanoid robots.”
He argues humanoid robots can:
- Do any human task
- Achieve lower mass-production cost than niche robots
Differing opinions among netizens:
- Pro humanoid dominance: AI speeds up R&D cycles dramatically.
- Pro specialization: Certain tasks require specialized hardware.
CEO Brett Adcock’s stance:
> “Humanoid robots will eventually dominate all forms globally.”
---
Reference Links
---
AI Content Tools for Robotics Creators
Platforms like AiToEarn官网 provide open-source, global AI content monetization solutions.
Features include:
- Multi-platform publishing (Douyin, Kwai, WeChat, Bilibili, Facebook, Instagram, YouTube, X/Twitter)
- Seamless AI content generation + analytics + model ranking (AI模型排名)
- Efficient monetization of AI creativity
Such tools complement humanoid robotics adoption by enabling creators to share and monetize innovations like Figure 03 on a global scale.
