Agility Robotics
First US pure-play humanoid robot company to go public — $2.5B SPAC merger, 65,000+ workplace hours logged, and a factory that can build 10,000 Digits a year.
1. Core Product / Service
Agility Robotics builds Digit, a bipedal humanoid robot designed for warehouse and logistics material handling. Digit stands roughly 5'9" tall, weighs ~140 lbs, and can lift up to 35 lbs — purpose-built for tote-moving tasks in existing warehouse workflows, not general-purpose manipulation.
Digit v5 (2026) is being positioned as the "first cooperatively safe humanoid robot" — capable of working alongside humans without safety cages, a critical unlock for warehouse deployment. The company also offers Arc, a software/cloud orchestration layer for managing fleets of Digits.
Agility operates RoboFab, a 70,000 sq ft manufacturing facility in Salem, Oregon, with peak capacity of 10,000 robots per year — the world's first full-scale humanoid factory [1].
2. Target Users & Pain Points
Agility targets warehouse and distribution center operators facing chronic labor shortages in material handling roles. The core pain point: tote-moving (picking items from shelves into totes, moving totes between stations) is repetitive, physically demanding, and has high turnover — exactly the profile of tasks that Digit addresses.
Current customers include Amazon, GXO Logistics, Toyota Motor Manufacturing Canada, Mercado Libre, and Schaeffler. Agility has logged over 65,000 hours of real-world operation and moved over 100,000 totes at a single Georgia facility [1]. Unlike general-purpose humanoid plays, Agility is narrowly focused on the $1T US market for material handling labor in manufacturing, distribution, and logistics.
3. Competitive Landscape
| Competitor | Focus | Scale |
|---|---|---|
| Figure AI | General-purpose humanoid for home/industry | $3.68B raised, $39B valuation |
| Tesla Optimus | Humanoid for manufacturing | Tesla-scale resources |
| Unitree | Low-cost humanoid + quadruped | $252M revenue, $6.2B IPO |
| 1X (Norway) | Wheeled humanoid for indoor use | OpenAI-backed |
Agility's key differentiator: proven warehouse deployment at scale. With 100 robots across 9 customer sites, $300M+ in multi-year orders for Digit v5, and a narrow logistics focus rather than general-purpose ambition, Agility is furthest along in real-world commercial deployment among Western humanoid companies [2].
4. Unique Observations
- The "pragmatic humanoid" play. Agility explicitly avoids the "robot in every home" narrative. Digit is a warehouse tool first — bipedal because warehouses are built for humans, but task-scoped to tote-moving. This narrow focus may be exactly what makes the economics work in the near term.
- NVIDIA Halos safety partnership as moat. NVIDIA selected Agility as the first partner for its Halos for Robotics safety system. Cooperative safety — working cage-free alongside humans — is a regulatory and technical barrier that Agility is positioned to clear first, potentially creating a 12–18 month regulatory-moat window [2].
- SPAC as a test case for the sector. Agility's June 2026 merger with Churchill Capital Corp XI at $2.5B (only 19% premium over its last private round) is the public market's first chance to price a pure-play humanoid company. If AGLT trades well, it opens the door for Figure, 1X, and others. If it doesn't, the entire sector's IPO window narrows [2].
5. Financials / Funding
- SPAC merger: Churchill Capital Corp XI (June 2026), valuation ~$2.5B, gross proceeds >$620M ($420M trust + ~$200M PIPE), PIPE led by Foxconn [2]
- Total raised privately: ~$640–652M across multiple rounds
- Last private round: $400M Series C (March 2025) at ~$2.1B post-money [2]
- Key investors: Amazon (Industrial Innovation Fund), NVIDIA, SoftBank Vision Fund 2, Foxconn, DCVC
- Orders: $300M+ in multi-year orders for Digit v5 [2]
- Pricing: Earlier Digit units estimated at ~$250,000 each
6. People & Relationships
- CEO: Peggy Johnson (former Microsoft EVP of Business Development, former Magic Leap CEO)
- Co-Founder & Chief Robot Officer: Jonathan Hurst (Oregon State University professor)
- Co-Founder & Chairman: Damion Shelton
- CTO: Pras Velagapudi
- CBO: Daniel Diez
- Partners: NVIDIA (Halos safety), Google DeepMind, Manhattan Associates, AWS
- Customers: Amazon, GXO, Toyota Motor Manufacturing Canada, Mercado Libre, Schaeffler
- Competitors: Figure AI, Tesla Optimus, unitree, 1X
Agility Robotics(Digit)
Digit 双足仓储机器人,客户现场运行时长是其主要数据叙事
- 数据类型
- 部署运行数据
- 规模口径
- 累计 65,000+ 运行小时(公司口径,二手转引)
- 客户
- Schaeffler;GXO;Toyota Motor Manufacturing Canada;Mercado Libre
做什么
Agility 的 Digit 是面向仓储物流搬运料箱的双足机器人。公开报道列出的客户包括 Schaeffler、GXO、Toyota Motor Manufacturing Canada 与 Mercado Libre,另有报道称 Amazon 曾试用 [1][2]。
65K 小时说法
"累计 65,000 多运行小时"为公司口径,CEO Peggy Johnson 亦引用同一数字;仅一份二手分析称其来自 9 个客户站点,并称出自 S-4 文件,本次未能核对原文,均标 ⚠️ [1][2]。公开材料未给出分客户小时数,也无第三方审计。
数据回流
公开信息中,这些小时主要被用作可靠性与商业化证据,未见公司披露客户现场数据如何回流训练(如遥操接管、失败片段或 RL 微调)及其规模,因此不能把运行小时等同于训练数据小时 ⚠️。对比 figure-ai 的 BMW 部署同样只披露运行时长。部署回流的方法论见 deployment-rl。
口径
SPAC 估值与订单结构来自二手报道,需以 S-4 原文为准 ⚠️。
全身控制靠仿真(补)
官方文章称其全身控制基础模型是小于 1M 参数的 LSTM,完全在 Isaac Sim/Lab 训练、零样本迁移到真机,客户 CAD/BIM 可导入仿真场景;外部数据供应方仅 NVIDIA 工具栈 [3],见 isaac-sim-lab。数据叙事因此以 L3 仿真为主,运行小时数用于可靠性证据。
- https://www.geekwire.com/2026/digit-maker-agility-robotics-to-go-public-in-2-5b-deal-heres-what-the-filings-say-about-its-finances/ (2026-10-14,经搜索摘要转引)
- https://businessmodelanalyst.com/agility-robotics-spac-raas-order-book/ (2026-10-14,二手)
- https://www.agilityrobotics.com/content/training-a-whole-body-control-foundation-model (2026-10-14)