휴머노이드 로봇용 배터리 : 세계 시장(2027-2040년)

Battery runtime has become one of the main limits on how far and how fast humanoid robots can scale. Most humanoids run for two to four hours per charge, against working shifts of eight to twelve hours. Adding cells is not a simple fix, because a heavier pack loads the same actuators it powers, and the battery has to fit inside a compact torso. Humanoid batteries must combine four things that rarely come together: high energy density, sharp peak power for walking and lifting, long cycle life across multiple shifts, and safety close to people. Robot makers including Figure AI and Tesla have concluded that off-the-shelf EV packs are a poor fit, and design their own packs around the robot's body and duty cycle.

Lithium-ion carries the market today:

• High-nickel NMC dominates where energy density matters most.
• LFP serves lower-cost and lower-endurance robots.
• Cylindrical cells lead on cost, maturity and supply. Pouch and custom-shaped cells are used where space is irregular.
• High-rate power cells, such as Molicel's, handle actuator peak loads.

Operators are tackling the runtime gap through system design as much as chemistry. Battery swapping is spreading in industrial settings: Hexagon's AEON and UBTECH's Walker S2 change their own batteries, and Apptronik and others use dual-battery designs. Docking and fast charging suit service and home robots. Pack architecture, thermal management and battery management software are becoming points of competition for robot makers. Next-generation chemistries are moving forward in parallel. These include silicon anodes from companies such as Amprius and Enovix, semi-solid and all-solid-state cells, lithium-metal and lithium-sulfur. Robots are attractive early customers for these premium cells because a pack of around 2 kWh can absorb a higher cost per kWh than a 60–100 kWh EV pack. Research groups argue that full-shift operation will ultimately need cells above 1,000 Wh/L. The supply base is global and broadening:

• Asian cell makers in Japan, Korea and China are developing robot-specific products. Panasonic, for example, is designing cylindrical cells for humanoids.
• European specialists include Saft, which partners with Enchanted Tools, and RRC Power Solutions.
• US developers QuantumScape, Factorial and others are positioning next-generation cells for robotics.

Batteries for Humanoid Robots: Global Market 2027-2040 analyzes the battery technologies, pack designs, charging strategies and supply chains that will determine how quickly humanoid robots move from pilot projects to full-shift deployment. With most humanoids limited to two to four hours of runtime, the battery is now central to the economics of every robot. Cell makers, pack integrators and robot makers across Asia, North America and Europe are developing robot-specific solutions. The report sets out what a humanoid battery must do: energy density in a confined torso, peak power for dynamic motion, multi-shift cycle life, safety near people, and a cost share of about 10% of the robot price.

Report contents include:

• Executive summary with key findings, the technology roadmap, and demand and value forecasts
• Humanoid market context: shipments, installed base, sector penetration, regional markets and government programmes
• The battery's share of the robot's bill of materials and conditions for a sub-US$20,000 humanoid
• Power demand by subsystem and activity, and energy needs per shift
• Battery requirements, a gap analysis for 8-hour operation, and specifications of leading humanoids
• Lithium-ion chemistries, cathodes, anodes and high-rate power cells
• Semi-solid, all-solid-state, lithium-metal, lithium-sulfur and sodium-ion batteries, with commercialization timelines
• Cell formats, pack sizing, pack architectures, BMS, thermal management and safety
• Fast charging versus battery swapping, total cost of ownership and the swap-pack multiplier
• Supply chain, supplier tiers, regional analysis across Asia, North America and Europe, automaker programmes and partnerships
• End-use applications: industrial, logistics, service, healthcare, household and hazardous environments
• Forecasts for 2027–2040 by application, chemistry, cell format and region
• Standards, regulation, barriers to scale, SWOT analysis and outlook to 2040
• 82 company profiles. Companies profiled include 24M Technologies, Ampace, Ampcera, Amprius Technologies, ATL (Amperex Technology Ltd.), Azure, BAK Power Battery, Basquevolt, Blue Current, Blue Solutions, BTR New Material Group, BYD / FinDreams Battery, CALB, CATL, Chongqing Tailan New Energy, CNS Battery, Echion Technologies, EcoPro BM, Enovix, Enpower, EVE Energy, Factorial Energy, Farasis Energy (Funeng Technology), Ganfeng Lithium, Gotion High-Tech, Great Power Energy, Grepow, Group14 Technologies, Hoppt Battery, Idemitsu Kosan, IL Science, Ilika, Ion Storage Systems, Joyson New Energy, LG Energy Solution, Lishen Battery, Lyten, Maxell, Microvast, Molicel (E-One Moli Energy), Murata Manufacturing, Nandu Power, Nyobolt, Panasonic Energy and more.....

Table of Contents

1 EXECUTIVE SUMMARY

• 1.1 Report scope and key definitions
• 1.2 Why the battery is the binding constraint on humanoid deployment
• 1.3 The humanoid market in 2026: hype vs. deployment reality
• 1.4 Humanoid vs. EV battery requirements
• 1.5 Battery chemistry roadmap 2027–2040
• 1.6 Solid-state batteries: why humanoids are solid-state's first commercial market
• 1.7 Market reality check: a premium niche of ~0.1 GWh in
• 1.8 Charging vs. battery swapping: implications for battery demand
• 1.9 Competitive landscape
• 1.10 Global humanoid battery demand 2027–2040 (GWh)
• 1.11 Global humanoid battery market value 2027–2040 (US$ million)
• 1.12 Comparison of third-party forecasts
• 1.13 Market drivers and restraints
• 1.14 Strategic recommendations

2 HUMANOID ROBOT MARKET CONTEXT AND POWER DEMAND

• 2.1 Industry overview
  • 2.1.1 From demonstration to pilot deployment
  • 2.1.2 The industrial robot baseline: five million robots in factories
  • 2.1.3 Humanoid sales 2025: ~7,000 units for industrial and professional use
  • 2.1.4 2026 shipments
  • 2.1.5 Robot battery demand 2025–2026
  • 2.1.6 Capital markets
  • 2.1.7 Specialised physical AI vs. humanoids: where value is captured
• 2.2 Humanoid shipment and installed base outlook 2027–2040
  • 2.2.1 Penetration by sector
• 2.3 Regional markets
  • 2.3.1 China
  • 2.3.2 United States
  • 2.3.3 South Korea
  • 2.3.4 Japan
  • 2.3.5 Europe
  • 2.3.6 Rest of World
• 2.4 Government policy and industrial strategy
• 2.5 Humanoid cost structure and the battery share of BOM
  • 2.5.1 Humanoid BOM breakdown
  • 2.5.2 Battery cost share of robot price and the ~10% industry target
  • 2.5.3 Conditions for a sub-US$20,000 humanoid and implications for battery cost
• 2.6 Power demand of a humanoid robot
  • 2.6.1 What the battery powers
  • 2.6.2 Power consumption by subsystem and activity
  • 2.6.3 Human caloric benchmark and cost of transport
  • 2.6.4 Energy demand of example applications, including 10+ kWh use cases
  • 2.6.5 AI compute: a growing share of the energy budget

3 BATTERY REQUIREMENTS AND DESIGN CONSTRAINTS

• 3.1 Requirements compared with other mobile platforms
• 3.2 Energy density: gravimetric and volumetric
  • 3.2.1 Torso volume and mass limits: packs confined to the chest compartment
  • 3.2.2 Volumetric vs. gravimetric energy density: why Wh/L matters more than in EVs
  • 3.2.3 The weight paradox: extra battery mass raises actuator load
  • 3.2.4 Target: >1,000 Wh/L cells for full-shift operation
• 3.3 Power density and peak discharge for dynamic motion
• 3.4 Cycle life and calendar life under multi-shift duty
  • 3.4.1 Cycling profile of humanoid robots
• 3.5 Safety for close human interaction
  • 3.5.1 Thermal runaway, nail penetration and fire containment
  • 3.5.2 Fall, impact, vibration and shock from bipedal locomotion
• 3.6 Operating temperature range
• 3.7 Fast-charge capability
• 3.8 Weight distribution, centre of mass and balance
• 3.9 Current state of humanoid batteries
  • 3.9.1 Specifications of commercial humanoids
  • 3.9.2 Bipedal (<1 kWh typical) vs. wheeled (>1.5 kWh) humanoids
  • 3.9.3 Average capacity per robot 2025–2040
• 3.10 Gap analysis: the road to 8-hour operation
  • 3.10.1 Runtime today: ~2 hours for transitional packs
  • 3.10.2 The 8-hour shift threshold, expected around
• 3.11 Designing the robot around the pack: battery-led robot architecture
• 3.12 Key battery challenges

4 BATTERY CHEMISTRIES AND CELL TECHNOLOGIES

• 4.1 Selection criteria and overview
• 4.2 Lithium-ion batteries
  • 4.2.1 Cathodes: NMC, NCA, LFP, LMFP, LCO, LMO
  • 4.2.2 High-nickel NMC (NMC622, NMC811, NMC9½½): the current incumbent
  • 4.2.3 LFP: cost advantage in lower-endurance and voice-interaction robots
  • 4.2.4 Anodes: graphite, silicon-carbon and silicon-dominant
  • 4.2.5 High-rate power cells for actuator peak loads
  • 4.2.6 Lithium-polymer packs
• 4.3 Semi-solid and quasi-solid-state batteries
  • 4.3.1 Technology and commercial status
  • 4.3.2 Humanoid deployments
• 4.4 All-solid-state batteries
  • 4.4.1 Sulfide, oxide, polymer and halide electrolytes
  • 4.4.2 Electrode technologies: high-nickel cathodes, silicon, lithium metal, anode-free
  • 4.4.3 Why humanoids commercialize solid-state before EVs: <2 kWh robot packs vs. 60–100 kWh EV packs
  • 4.4.4 Robots as the premium-tolerant customer: energy density, peak power and safety
  • 4.4.5 Humanoid solid-state programmes
    • 4.4.5.1 XPENG IRON (2nd generation)
    • 4.4.5.2 GAC GoMate (3rd generation)
    • 4.4.5.3 EngineAI (Zhongqing) T800
    • 4.4.5.4 Farasis (Funeng Technology)
    • 4.4.5.5 CALB
    • 4.4.5.6 Joyson–Enpower embodied-AI battery joint venture
    • 4.4.5.7 EVE Energy Longquan No.
  • 4.4.6 Cost: US$400–600/kWh today vs. LFP and NCM packs
  • 4.4.7 Samsung SDI's two-track strategy: cylindrical for mass-market robots, all-solid-state for premium humanoids
    • 4.4.7.1 Pouch all-solid-state for physical AI vs. prismatic for EVs
    • 4.4.7.2 S-Line pilot: sulfide electrolyte, anode-free design, customer evaluation
  • 4.4.8 Mass production plans of key solid-state players
  • 4.4.9 Unconverged routes: oxide, sulfide and polymer as parallel chemistry bets
  • 4.4.10 Automotive OEM solid-state timelines and spill-over to humanoids
  • 4.4.11 Robots as the field-data generator for automotive solid-state programmes
• 4.5 Lithium-metal and anode-free batteries
• 4.6 Lithium-sulfur batteries
  • 4.6.1 Advantages and challenges
  • 4.6.2 Humanoid applications
• 4.7 Sodium-ion batteries: low-cost household robots and docking stations
• 4.8 Long-term and complementary technologies
  • 4.8.1 Metal-air (Li-air, Zn-air)
  • 4.8.2 Structural batteries (EFSB) for limbs and tendons
  • 4.8.3 Supercapacitors and hybrid storage for peak power
  • 4.8.4 Fuel cells for long-endurance robots
• 4.9 Chemistry adoption roadmap 2027–2040

5 CELL FORMATS, PACK INTEGRATION, BMS, THERMAL MANAGEMENT AND SAFETY

• 5.1 Cell formats
  • 5.1.1 Cylindrical: 18650, 21700, 46-series
  • 5.1.2 Pouch: Samsung SDI's shift to pouch format for robots
  • 5.1.3 Prismatic
  • 5.1.4 Structural and custom-shaped cells
• 5.2 Cell count and pack sizing
  • 5.2.1 21700-based pack estimation
  • 5.2.2 4680-based pack estimation
  • 5.2.3 Installed vs. actual capacity: fit assessment for major models
• 5.3 Pack architectures
  • 5.3.1 Centralised torso pack
  • 5.3.2 Distributed and embedded packs
  • 5.3.3 Swappable cartridges: single and dual
  • 5.3.4 Structural integration: torso as battery enclosure
  • 5.3.5 Per-model pack customisation vs. platform standardisation
• 5.4 Voltage platforms and power distribution
• 5.5 Battery management systems
  • 5.5.1 SoC/SoH estimation under dynamic loads
  • 5.5.2 AI-driven BMS and predictive energy management
  • 5.5.3 Functional safety and robot-controller integration
• 5.6 Thermal management
  • 5.6.1 Air, heat-pipe and liquid cooling
  • 5.6.2 Shared cooling of battery and AI compute
  • 5.6.3 Case studies
  • 5.6.4 Figure 03 battery design
  • 5.6.5 Tesla Optimus: pack, cells and patent PCT/US2023/033983
  • 5.6.6 Boston Dynamics Atlas battery strategy
  • 5.6.7 UBTECH Walker S2 dual-battery system
  • 5.6.8 High-power pack for multi-legged robots (Molicel P45B 16S3P)
• 5.7 Connectors, harnesses and protection circuits

6 CHARGING, BATTERY SWAPPING AND ENERGY MANAGEMENT

• 6.1 Limited endurance: fast charging or swapping?
• 6.2 Fast charging (5C–7C) and degradation trade-offs
• 6.3 Docking and autonomous charging
• 6.4 Battery swapping
  • 6.4.1 Manual hot swapping
  • 6.4.2 Autonomous self-swapping
  • 6.4.3 Dual-battery architectures
  • 6.4.4 Swap stations and fleet energy management
  • 6.4.5 Standardisation of swappable packs
• 6.5 Economics of battery swapping
  • 6.5.1 Additional packs per robot and effect on total battery demand
• 6.6 Charging model by application
• 6.7 Wireless power transfer
• 6.8 Energy harvesting and regenerative systems

7 SUPPLY CHAIN AND COMPETITIVE LANDSCAPE

• 7.1 Humanoid battery value chain
• 7.2 Market scale today: strategic, not yet material
  • 7.2.1 Robot battery shipments vs. total Li-ion (5.2 GWh in 2024, <0.4%)
  • 7.2.2 Humanoid battery demand of ~0.1 GWh in 2026 vs. gigafactory scale
  • 7.2.3 Battery vendors' narratives vs. robot OEM priorities
• 7.3 Supplier–OEM relationships
• 7.4 Regional analysis
  • 7.4.1 China: scale, first-mover solid-state programmes and a full domestic supply chain
  • 7.4.2 South Korea: EV-to-robot pivot
  • 7.4.3 Japan: Panasonic, TDK, Murata, Maxell
  • 7.4.4 Taiwan, US and Europe
• 7.5 Supplier tiers
• 7.6 Automotive OEMs as humanoid makers and battery integrators
  • 7.6.1 XPENG Guangzhou humanoid production base
  • 7.6.2 GAC GoMate small-batch production
• 7.7 Materials suppliers
• 7.8 Specialist robotics pack integrators
• 7.9 Investments, partnerships, joint ventures and M&A 2024–2026
• 7.10 Humanoids in battery manufacturing: CATL, BMW and the reverse relationship

8 END-USE APPLICATIONS

• 8.1 Industrial manufacturing
  • 8.1.1 Automotive and battery plants
  • 8.1.2 Electronics and general manufacturing
• 8.2 Logistics and warehousing
• 8.3 Commercial and service
  • 8.3.1 Retail and hospitality
  • 8.3.2 Healthcare and eldercare (KAPEX hospital deployment)
• 8.4 Household and domestic
• 8.5 Research, education and AI training-data collection
• 8.6 Hazardous environments, defence and outdoor
• 8.7 Battery requirements compared across applications

9 GLOBAL MARKET FORECASTS 2027–2040

• 9.1 Forecast summary
• 9.2 Scenarios and assumptions
• 9.3 Battery demand (GWh)
  • 9.3.1 By application
  • 9.3.2 Industrial demand by technology
  • 9.3.3 Industrial demand including battery swapping
  • 9.3.4 Commercial and domestic demand by technology
  • 9.3.5 By chemistry
  • 9.3.6 Solid-state battery demand from humanoids
  • 9.3.7 Lithium-ion NMC by nickel content
  • 9.3.8 By cell format
  • 9.3.9 By region
• 9.4 Market value (US$ million)
  • 9.4.1 Average selling prices
  • 9.4.2 By application
  • 9.4.3 By chemistry
  • 9.4.4 Solid-state batteries by application
  • 9.4.5 By region
  • 9.4.6 Value split: cells, packs, BMS and thermal systems
• 9.5 Scenario analysis
• 9.6 Sensitivity analysis

10 REGULATION, STANDARDS, CHALLENGES AND OUTLOOK

• 10.1 Safety standards and certification
• 10.2 Standards gap: no robot battery duty-cycle or performance test
• 10.3 China's solid-state definitions: GB/T 43568-2026 and the IEC submission
• 10.4 EU Battery Regulation: removability, battery passport and recycling
• 10.5 Transport, end-of-life and second life
• 10.6 Supply chain risks: critical minerals, export controls and tariffs
• 10.7 Technical challenges
• 10.8 Market challenges: undefined product-market fit, fast-changing robot architectures, supplier hype
  • 10.8.1 Robot body designs iterating faster than battery qualification cycles
  • 10.8.2 Lack of pack standardisation across platforms
  • 10.8.3 Solid-state chemistry routes yet to converge
• 10.9 SWOT analysis
• 10.10 Outlook to

11 COMPANY PROFILES (82 company profiles)

12 APPENDICES

• 12.1 Report scope and objectives
• 12.2 Definitions and terminology
• 12.3 Humanoid robot classification
  • 12.3.1 Bipedal, wheeled and hybrid humanoids
  • 12.3.2 Autonomy levels: transition from Level 3 to Level
• 12.4 Research methodology
  • 12.4.1 Primary and secondary research
  • 12.4.2 Forecast model
  • 12.4.3 Scenarios: base, conservative and accelerated
• 12.5 Market segmentation
• 12.6 Glossary of terms
• 12.7 List of abbreviations

13 REFERENCES