Figure 03 enables cable-free charging via wireless power

PLUS: Microsoft’s industrial humanoid and 1X NEO launching


Figure 03 enables cable-free charging via wireless power

Welcome back to your Robot Briefing

Figure's latest humanoid can now walk up to a charging pad and power itself back up without any cables or human help. The company built a wireless charging system from scratch because nothing on the market could handle the power requirements for a full-size humanoid that needs to work multiple shifts.

As more robotics companies chase deployment at scale, the infrastructure around the robot — charging, fleet management, maintenance — may matter as much as the robot itself. Can companies justify humanoid investments if each unit still needs dedicated charging stations and custom facility modifications?

In today's Robot update:

Figure 03 charges wirelessly through its feet
Microsoft backs industrial humanoid for inspection tasks
ByteDance solves the robot hand training data problem
1X's NEO humanoid ships at $20,000
News

Figure 03 Cuts the Cord with Wireless Charging

Snapshot: Figure has equipped its Figure 03 humanoid robot with wireless inductive charging that lets the robot autonomously recharge by stepping onto a charging pad, eliminating cables and enabling continuous operation throughout the day.

Breakdown:

The robot charges at 2kW through coils embedded in its feet, a capability Figure engineered from scratch because no existing solution met their requirements for humanoid robotics.
In home or facility settings, Figure 03 can autonomously dock and recharge as needed without human intervention, enabling it to manage its own power throughout multi-hour deployments.
The custom-built charging infrastructure signals Figure's focus on deployment readiness rather than just demonstration capabilities, addressing one of the practical barriers to leaving robots unsupervised in real environments.

Takeaway: Infrastructure innovations like autonomous charging are becoming key differentiators that separate lab demos from deployment-ready systems. Companies evaluating humanoid robots should now assess not just task capability but the complete operational ecosystem required for autonomous multi-shift operation.

News

Microsoft & Hexagon Deploy Industrial Humanoid 'AEON'

Snapshot: Microsoft has joined forces with Hexagon Robotics to deploy AEON, an industrial humanoid robot designed for inspection and maintenance tasks, powered by Azure AI and targeting automotive, aerospace, and manufacturing sectors. The partnership announcement signals that cloud giants now view industrial humanoids as commercially viable, not just research projects.

Breakdown:

AEON targets specific industrial tasks like defect detection and equipment inspection rather than general-purpose factory work, letting companies deploy humanoids for high-value use cases without rebuilding entire facilities.
The robot combines Hexagon's sensor fusion technology with Microsoft's Azure cloud infrastructure, meaning companies can manage robot fleets the same way they manage enterprise software with centralized training, updates, and monitoring.
Initial deployments focus on industries facing acute labour shortages in manual inspection roles, where humanoids can handle night shifts and hazardous environments while keeping humans in supervisory positions.

Takeaway: Microsoft's direct involvement moves industrial humanoids from the "interesting concept" category into the "evaluate for 2026-2027 deployment" timeline for mid-sized manufacturers. Companies in automotive, aerospace, and logistics now have a clear enterprise-grade option backed by familiar cloud infrastructure rather than betting on startup robotics vendors.

News

ByteDance Cracks the Dexterity Data Problem

Snapshot: ByteDance researchers partnered with MANUS to build ByteDexter, a system that uses high-fidelity data gloves to train robots on complex manipulation tasks, boosting success rates from 45% to 85% on previously unseen objects.

Breakdown:

ByteDexter addresses the core bottleneck in training dexterous robots: collecting enough high-quality demonstration data for systems with 56 total degrees of freedom (21 per hand plus dual arms) where traditional teleoperation methods fail.
The system combines MANUS data gloves to capture detailed finger movements with Meta Quest headsets for wrist tracking, allowing human operators to control two Franka robotic arms simultaneously during long-duration tasks while preserving the fine-grained hand kinematics needed for policy training.
On real-world benchmarks like makeup decluttering with unfamiliar table layouts, the trained system achieved an 85% success rate versus 45% for baseline approaches, and it can execute abstract commands like "pick up the drinkable object" even when that specific item wasn't in the training set.

Takeaway: This tackles the expensive, time-consuming problem of generating training data for dexterous robots—the kind of barrier that's kept human-level manipulation in research labs rather than factory floors. Companies evaluating robotic manipulation should watch whether this data collection approach gets adopted more broadly, as it could accelerate the timeline for deploying capable systems from years to months.

News

Domestic robots are finally coming home in 2026

Domestic robots are finally coming home in 2026

Image Source: Gemini / There's A Robot For That

Snapshot: 1X's NEO humanoid is launching to customers this year at $20,000, marking the first wave of multi-purpose domestic robots actually shipping to homes—finally putting a price and timeline on technology that's been promised for decades.

Breakdown:

NEO targets wealthy early adopters at $20,000 upfront or $500 monthly, with 1X explicitly planning to use these first customers as training data to improve the system's capabilities.
The promotional videos don't show the reality: human operators wearing VR headsets still remotely control these robots when they get confused, meaning true autonomy remains a work in progress even as products ship.
Silicon Valley startups backed by Nvidia and Jeff Bezos are leading the charge over tech giants, while China's government warns of a potential bubble as its domestic humanoid industry heats up.

Takeaway: This shifts the timeline from "someday" to "now" for companies with capital to deploy early-stage automation, but the heavy reliance on human teleoperation means this is still a pilot-and-learn phase rather than plug-and-play readiness. Smart operations leaders should track how these first deployments perform over the next 12-18 months before committing budgets.

Other Top Robot Stories

WIRobotics showcased its WIM S wearable walking-assist robot and ALLEX humanoid at CES 2026, drawing attention for having supplied wearable robots to consumers for three years before most competitors entered the market, with active discussions underway with NVIDIA, Meta, and Amazon regarding technical collaboration on humanoid applications.

Astute projects the global digital surgery technologies market will surge from $712.9 million in 2024 to $6.48 billion by 2033 at a 27.4% CAGR, driven by AI integration, robotics, and AR/VR technologies that are transforming surgical precision and training across healthcare systems worldwide.

Dyna presented its approach to general-purpose foundation models for physical tasks at CES 2026, with CEO Lindon Gao discussing how the company is scaling diverse manipulation capabilities to commercial deployment while their robot autonomously folded laundry in the background during the interview."

🤖 Your robotics thought for today:
Figure 03 autonomously docks itself for wireless charging throughout the day while most warehouse robots still need humans to swap batteries or plug in cables—so are we designing automation systems or just expensive supervised equipment?

Until tomorrow,
Uli

Figure 03 enables cable-free charging via wireless power

Great! Check your inbox and click the link to confirm.
Please enter a valid email address.