Agility Robotics

Agility Robotics is the American humanoid-robotics company founded in 2015 as an Oregon State University spinout, developer of the Digit commercial humanoid robot.
Agility Robotics

Agility Robotics is an American humanoid-robotics company headquartered in Albany, Oregon, with a dedicated manufacturing facility in Salem, Oregon, founded in 2015 as a commercial spinout from the Oregon State University Dynamic Robotics Laboratory by Damion Shelton, Jonathan Hurst, and Mikhail Jones. The company develops the Digit bipedal humanoid robot, the longest-running commercially deployed humanoid in the American market, and operates the only large-scale dedicated humanoid manufacturing facility (RoboFab in Salem) in the United States. As of 2026, Agility is one of the four principal American humanoid-robotics commercial players alongside Figure AI, Tesla AI, and Apptronik, with Amazon, GXO Logistics, and other warehouse-and-logistics customer deployments running continuously since 2023.

At a glance

  • Founded: 2015 in Albany, Oregon by Damion Shelton, Jonathan Hurst, and Mikhail Jones as a spinout from Oregon State University.
  • Status: Private.
  • Funding: Approximately $300 million cumulative through early 2026. Series B of $150 million in February 2024 led by DCVC and Playground Global, with previous-round participation from Amazon Industrial Innovation Fund, Sony Innovation Fund, and others.
  • CEO: Peggy Johnson (since March 2024). Former chief executive of Magic Leap and former Microsoft Business Development leader.
  • Other notable leadership: Jonathan Hurst (Co-founder and Chief Robot Officer; Oregon State University professor of mechanical engineering), Damion Shelton (Co-founder; prior CEO until 2024), Aindrea Campbell (Chief Operating Officer).
  • Open weights: None. Agility is a humanoid-hardware company with a proprietary AI control stack focused on locomotion, manipulation, and warehouse-task automation.
  • Flagship product: Digit bipedal humanoid robot, in commercial pilots since 2020 and in production deployment since 2023. Prior research platform: Cassie (2017 to 2019) bipedal robot, used as the locomotion-research platform that anchored the Digit design.

Origins

Agility Robotics was founded in 2015 as a commercial spinout from the Dynamic Robotics Laboratory at Oregon State University, where Jonathan Hurst had been a faculty member since 2008. The lab's research focused on bipedal-locomotion theory and on the design of compliant-leg robotic platforms; the founding thesis was that the locomotion-research lineage produced a commercially viable bipedal humanoid earlier than competitor approaches that started from upper-body manipulation.

The first generation of the company's hardware was the Cassie bipedal robot, released in 2017 as both a research platform and a commercial product for the small academic-and-research market. Cassie was an ostrich-like two-legged platform with no upper body, designed specifically for high-quality locomotion research; the platform was acquired by approximately a dozen research groups (including Carnegie Mellon, MIT CSAIL, and others) and accumulated several published demonstrations of bipedal locomotion at unprecedented speeds for an electric-actuated humanoid.

The transition from Cassie to Digit occurred from 2019 onward. Digit added a torso, arms, and an end-effector pair on top of the Cassie-derived bipedal platform, repositioning the company from a locomotion-research-hardware vendor to a commercial-humanoid platform vendor. Early Digit deployments at Ford in 2020 (last-mile package delivery research) and at adjacent industrial customers in 2021 and 2022 served as pilot-scale demonstration cycles that informed the production Digit specifications.

The Amazon warehouse pilots beginning in October 2023 were the company's commercial-scale customer engagement. Amazon had been an investor through the Industrial Innovation Fund since 2022, and the pilot deployment at the Amazon Sumner, Washington warehouse was the first announced warehouse deployment of a humanoid robot at production-adjacent scale globally. The pilots have continued through 2024, 2025, and 2026, with Agility describing the deployment as production-grade rather than pilot-grade by mid-2025.

Mission and strategy

Agility Robotics' stated mission is to "make robots that move you," operationalised through Digit as a general-purpose bipedal humanoid platform for industrial-task automation. The strategic premise is that warehouses and logistics centres are the highest-value initial humanoid deployment environment because the operational complexity is bounded enough for current AI control systems and the labour-shortage economics produce a clear customer pull at the unit-economics achievable.

The Salem, Oregon RoboFab manufacturing facility, opened in 2023, is the strategic anchor of the company's operating model. RoboFab is described publicly as the first dedicated humanoid-robot manufacturing facility globally, with a target production capacity of 10,000 humanoids per year at full ramp-up. The capacity claim is significant: at the scale targeted, RoboFab alone would exceed the total combined annual humanoid production of the rest of the US cohort. Whether the facility actually ramps to the stated capacity (and at what unit cost) is the most material commercial-and-operational signal for the company over the next 18 months.

The vertical positioning is in the middle of the humanoid market. Agility builds the hardware itself and develops the AI control stack in-house, but does not pursue general-purpose foundation-model leadership in the way that Figure AI has been pushing with Helix. The strategic bet is that warehouse-and-logistics tasks are bounded enough that vertical AI specialisation (warehouse-specific perception, picking, transport, charging cycle management) produces better unit economics than general-purpose foundation-model approaches. The bet's success will be visible in 2026 and 2027 deployment-side performance data.

Models and products

  • Digit (commercial bipedal humanoid, production deployment since 2023). Five-foot-nine-inch bipedal humanoid weighing approximately 140 pounds, designed for warehouse and logistics deployment. Two-arm manipulation with palm-and-grip end-effectors specifically designed for tote and box handling rather than fine-motor manipulation. Battery operation with autonomous-charge-cycle architecture (Digit walks to a charging station and continues operation in shifts). Computer-vision and inertial perception stack; locomotion-and-manipulation control developed in-house.
  • Cassie (research bipedal platform, 2017 to 2019). Predecessor research platform for the bipedal locomotion architecture that underlies Digit. Sold to approximately a dozen academic robotics research groups; no longer commercially produced.
  • RoboFab manufacturing facility (Salem, Oregon, opened 2023). Dedicated humanoid-robot manufacturing facility with stated production capacity target of 10,000 humanoids per year. The first large-scale humanoid-specific factory globally.

Benchmarks and standing

Standing in the humanoid-robotics category is measured through named customer deployments, hours of operating time, and demonstrated industrial-task coverage. Agility's deployment evidence at the Amazon Sumner warehouse (continuously since October 2023), at GXO Logistics, and at adjacent industrial customers places it in the leading group of commercially-deployed humanoid platforms. The Salem manufacturing facility's production capacity, if it ramps as targeted, would put Agility in the leadership position on per-year production volume in the US market.

The company has not published unit-economics data (cost per Digit, cost per hour of operation, cost per task completed) at the resolution that would allow comparison against the broader humanoid cohort. Industry coverage has placed Digit's per-unit pricing in the $200,000 to $300,000 range and the per-hour-of-operation cost (amortised plus consumables) in the $10 to $20 range as of 2025 estimates, with both numbers expected to fall through 2026 and 2027 as production volume scales.

Leadership

  • Peggy Johnson (Chief Executive Officer, since March 2024). Former chief executive of Magic Leap (2020 to 2023), where she led the company's enterprise pivot after a consumer-AR strategy reset. Prior experience as Microsoft's Executive Vice President for Business Development. Recruited to Agility to lead the company's scale-from-pilot-to-production transition.
  • Jonathan Hurst (Co-founder, Chief Robot Officer). Oregon State University professor of mechanical engineering and the academic anchor for the company's locomotion-research lineage. Continues as the senior technical-research executive.
  • Damion Shelton (Co-founder, former CEO until 2024). Held the chief-executive role from founding through the March 2024 transition to Peggy Johnson. Remains involved as a co-founder and board member.
  • Mikhail Jones (Co-founder). Engineering co-founder.
  • Aindrea Campbell (Chief Operating Officer). Operations executive recruited from former Apple manufacturing-operations roles. Leads the manufacturing-scale organisation around the Salem RoboFab facility.

Funding and backers

Agility Robotics has raised approximately $300 million in cumulative private capital through early 2026.

  • Series A (multiple tranches, 2018 to 2022): Approximately $150 million cumulative across several rounds. Participants include DCVC, Playground Global, Amazon Industrial Innovation Fund, Sony Innovation Fund, and adjacent industrial-and-strategic investors.
  • Series B (February 2024): $150 million led by DCVC and Playground Global. The round funded the RoboFab manufacturing ramp and the Series-B-pricing valuation was reported in the $1.5 billion range, putting Agility at the lower end of the humanoid-robotics unicorn tier.

The Amazon-investor-and-customer relationship is structurally important. Amazon participates as a strategic investor (through the Industrial Innovation Fund) and as the largest disclosed customer (Sumner pilot, with subsequent disclosed expansions). The two roles create a different kind of capital relationship than the standard investor-only or customer-only pattern, with strategic implications for both the warehouse-deployment commercial trajectory and the exit-path scenarios that institutional investors model.

Industry position

Agility Robotics' industry position is as the longest-running commercially deployed humanoid platform in the US market, with the strongest production-manufacturing infrastructure and the most-established warehouse-customer-deployment-relationship in the cohort. Within the US cohort, Agility is the warehouse-and-logistics specialist; Figure AI is the industrial-manufacturing-and-foundation-model entrant; Tesla AI Optimus is the consumer-and-internal-deployment vertically-integrated player; 1X is the consumer-and-household entrant; Apptronik is the industrial-and-NASA-partnership entrant.

The Chinese humanoid cluster competes on cost and on manufacturing scale. Agility's defence against the Chinese cluster is the US-customer-relationship and US-manufacturing position; the Salem RoboFab facility is part of the structural argument that Agility can produce competitively against Chinese cost-of-manufacturing without depending on Chinese supply chains for the final assembly. The defence holds or weakens based on the actual unit-cost comparison through 2026 and 2027.

Competitive landscape

  • Figure AI (US): Industrial-manufacturing-focused with BMW partnership; vertically integrated AI-and-hardware stack. Higher capital raised, higher valuation; less production-deployment evidence than Digit.
  • Tesla AI Optimus (US): Internal deployment in Tesla factories. Highest potential scale through Tesla manufacturing infrastructure. External commercial launch not yet realised.
  • 1X (Norway / US): Consumer-and-household positioning (NEO, EVE) rather than industrial.
  • Apptronik (US): Industrial-manufacturing-focused with Mercedes-Benz partnership; partner foundation-model AI stack (Google DeepMind collaboration).
  • Boston Dynamics (US, owned by Hyundai): Atlas (humanoid, transitioned to all-electric in 2024) plus Spot and Stretch. Different commercial-deployment pattern (Spot and Stretch lead Atlas commercially).
  • Chinese humanoid cluster: Unitree, Agibot, UBTECH, XPeng Robotics, Xiaomi Robotics. Cost-and-volume competition; different export-control and customer-procurement dynamics.
  • European entrants: Neura Robotics (Germany), Hexagon Robotics (Sweden/Switzerland). Smaller scale and different vertical orientations.

Outlook

Open questions and watchable signals over the next 6 to 18 months:

  • RoboFab production ramp. The Salem facility's stated 10,000-per-year capacity is a leading signal for the company's commercial trajectory. Actual production volume achieved in 2026 will be the most informative measurement.
  • Amazon deployment expansion. The Sumner pilot has been the company's flagship customer relationship. Whether it expands to additional Amazon warehouses, additional task domains, or to a stated production-scale fleet count will indicate whether the warehouse-deployment thesis is converting to fleet-scale revenue.
  • Peggy Johnson's operator transition completion. The March 2024 CEO transition from Damion Shelton to Peggy Johnson was the company's most significant leadership change. Whether the production-and-scale operating model that Johnson is responsible for delivers the ramp targets through 2026 will define her tenure.
  • Per-unit cost disclosure or implication. The first credible per-unit-cost number from Agility (in customer pricing announcements, in S-1 filings if the company files publicly, or in third-party deployment-data leaks) will reset the public framing of the humanoid-robotics category's economics.
  • Series C or later round pricing. A subsequent funding round in 2026 or 2027 will reveal the institutional-capital view of Agility's strategic position relative to the broader cohort. The Series B at $1.5 billion was at the lower end of the unicorn tier; a Series C at materially higher valuation would suggest the deployment-and-production-evidence has shifted institutional belief.

Sources

About the author
Nextomoro

Nextomoro

nextomoro tracks progress for AI research labs, models, and what's next.

AI Research Lab Intelligence

nextomoro tracks progress for AI research labs, models, and what's next.

AI Research Lab Intelligence

Great! You’ve successfully signed up.

Welcome back! You've successfully signed in.

You've successfully subscribed to AI Research Lab Intelligence.

Success! Check your email for magic link to sign-in.

Success! Your billing info has been updated.

Your billing was not updated.