American Humanoid Robots

Overview: Optimus is Tesla’s humanoid robot program for unsafe, repetitive or boring work. It belongs in the American ecosystem because Tesla has manufacturing scale, AI inference hardware, vision software and factory environments where a humanoid could be tested.

Technical notes: Tesla’s official AI page frames Optimus around balance, navigation, perception and interaction with the physical world. The page should not invent payload, battery life, hand degrees of freedom or cost unless those numbers come from Tesla source material.

Mobility: Bipedal humanoid

Autonomy level: Development program. Public demonstrations, internal tests and future production claims should be separated from verified deployment.

Why it matters: Optimus matters because Tesla can connect robot hardware with vehicle-scale perception engineering, in-house manufacturing and a large real-world operations network. The open question is whether that stack becomes reliable enough for repeatable physical work.

Limits and current status: Treat Optimus as an active development platform. Do not present it as a proven household robot or broad commercial worker. Public demos may involve scripted behaviors, teleoperation, resets or limited task scope.

Overview: Figure 01 is the first visible platform from Figure AI. It set the design direction for an American startup focused on humanoids for manufacturing, logistics and warehousing.

Technical notes: Use Figure 01 to explain iteration. It showed the company’s first full-body approach before Figure 02 added a cleaner industrial design, more compute and a redesigned hand system.

Mobility: Bipedal humanoid

Autonomy level: Early development and demonstrations, not a broad deployment platform.

Why it matters: Figure 01 matters because it turned Figure from a funded startup into a visible hardware company. It also created the baseline for later BMW factory tests and Figure’s newer product line.

Limits and current status: Do not overstate Figure 01. The stronger current evidence sits around Figure 02, Figure 03 and Helix. Figure 01 belongs in the history section and comparison table, not as the latest platform.

Overview: Figure 02 is one of the most visible American industrial humanoids. It is designed around a human-scale body, hands, onboard perception and a cleaner hardware package than Figure 01.

Technical notes: Figure 02 is commonly described with six RGB cameras, onboard compute, integrated wiring, a torso battery and five-finger hands. The page should present those details as model-specific and source-dependent.

Mobility: Bipedal humanoid

Autonomy level: Industrial trials and demos. Claims about autonomy should be tied to the specific source and task.

Why it matters: Figure 02 matters because it moved Figure into factory discussion with BMW and gave the US market a polished industrial humanoid competitor next to Tesla, Apptronik, Agility and Boston Dynamics.

Limits and current status: A factory trial is not the same as large-scale deployment. The article should separate demonstration footage, customer testing and repeatable commercial work.

Overview: Figure 03 is included because the live Figure site now positions it as the current general-purpose humanoid robot. It should sit above Figure 02 in recency while Figure 02 remains important for factory context.

Technical notes: The key technical story is the coupling of Figure hardware with Helix. The page should avoid claiming household autonomy across all chores until independent deployment evidence exists.

Mobility: Bipedal humanoid

Autonomy level: Figure says Helix supports navigation in changing environments. Treat home capability as early and environment-dependent.

Why it matters: Figure 03 shows how American humanoid startups are moving from factory pilots toward home-facing narratives. That shift increases the need for careful language about teleoperation, privacy and task limits.

Limits and current status: Do not call Figure 03 a solved home robot. Real homes include clutter, pets, fragile objects, stairs, private data and unpredictable human behavior.

Overview: Digit is one of the clearest American attempts to move humanoid robots from lab demonstrations into warehouse labor. Its body is not a human copy. The legs, torso and arms are optimized for moving totes in spaces built for people.

Technical notes: Agility describes a progression from Oregon State research to Cassie and then Digit. The current public company page lists Salem, Oregon, RoboFab and industrial customers while emphasizing real deployment over general claims.

Mobility: Bipedal humanoid with arms and non-humanlike leg geometry

Autonomy level: Industrial deployment in constrained workflows. Human-safe collaboration and broader autonomy remain staged by certification and environment.

Why it matters: Digit matters because it is tied to logistics customers, a dedicated Oregon factory and a narrow material-handling job. This makes it more measurable than a broad home-assistant promise.

Limits and current status: Digit’s strongest evidence sits in structured warehouse workflows. Do not present it as a general household humanoid or a robot that can replace every warehouse task.

Overview: Cassie should be included with a clear label. It is not a full humanoid worker. It is the bipedal foundation that helped Agility move toward Digit.

Technical notes: Cassie’s value is locomotion. The platform removes human-like upper-body complexity so researchers can focus on dynamic walking, control and balance.

Mobility: Bipedal legs without a full humanoid upper body

Autonomy level: Research locomotion platform.

Why it matters: Cassie matters because the US humanoid ecosystem depends on legged locomotion research, not only hands and AI models. It is a bridge from university lab work to industrial humanoid design.

Limits and current status: Do not classify Cassie as a full humanoid robot. It belongs in research and history sections, not in the active industrial humanoid table without a warning.

Overview: Apollo is Apptronik’s Austin-built humanoid platform for logistics and industrial work. It comes from a company with experience building more than ten previous robots, including work connected to NASA Valkyrie.

Technical notes: Apptronik lists Apollo as 5 ft 8 in tall, 160 lb, with a 55 lb payload and four hours per battery pack. The design is modular and can be stationary or mobile with legs.

Mobility: Modular humanoid, stationary or mobile with legs depending on configuration

Autonomy level: Customer pilots and training environments. Apptronik emphasizes software integration and point-and-click fleet control.

Why it matters: Apollo matters because it ties Texas humanoid robotics to NASA-linked engineering, Google DeepMind collaboration, Mercedes-Benz interest and industrial data collection.

Limits and current status: Apollo should not be written as deployed at broad commercial scale. It belongs in pilots, training environments and customer evaluation unless a source confirms production deployment.

Overview: The electric Atlas is the post-hydraulic generation of Boston Dynamics’ humanoid work. It should be separated from the older Atlas because the platform, commercial aim and hardware package changed.

Technical notes: Boston Dynamics now presents Atlas as an enterprise humanoid with high-power actuation, 56 DoF, tactile sensing, 360-degree camera view and self-swappable batteries. These specs should be attached directly to the official Atlas page.

Mobility: Bipedal humanoid with high range of motion

Autonomy level: Boston Dynamics describes autonomous work with minimal supervision for material handling. State task scope clearly.

Why it matters: Atlas matters because Boston Dynamics has spent more than a decade testing dynamic whole-body movement. The electric platform moves that research closer to industrial work instead of viral motion demos alone.

Limits and current status: Even strong autonomous task videos do not prove broad deployment. Keep the wording tied to part handling, material handling and named pilots.

Overview: The hydraulic Atlas made Boston Dynamics a global reference for humanoid movement. It ran, jumped, flipped and fell in videos that made dynamic humanoid control visible to the public.

Technical notes: This older platform should be framed as hydraulic research hardware, not the current enterprise product. It belongs in the history and DARPA sections.

Mobility: Hydraulic bipedal humanoid

Autonomy level: Historical research and demonstrations.

Why it matters: Hydraulic Atlas matters because many current humanoid expectations trace back to Boston Dynamics proving that a human-sized robot could perform aggressive whole-body motion.

Limits and current status: The platform is historical. Do not use old parkour videos as proof that the new electric Atlas is commercially deployed everywhere.

Overview: PETMAN belongs in the page because it is an American humanoid ancestor to Atlas. Its job was not warehouse work. It tested protective clothing under human-like motion.

Technical notes: The technical value is repeatable human-like movement in controlled test conditions, especially for protective suit evaluation.

Mobility: Bipedal humanoid mannequin-like robot

Autonomy level: Historical test platform.

Why it matters: PETMAN shows that US humanoid robotics was not only about social robots or AI. Some early platforms were built for military, safety and human-equipment testing.

Limits and current status: PETMAN is historical. Do not classify it as an active humanoid company product.

Overview: Valkyrie, also called R5, was built by NASA Johnson Space Center for disaster-response and space-related robotics research. It connects humanoid robotics to remote work where humans may be delayed, distant or at risk.

Technical notes: Valkyrie profiles should focus on whole-body mobility, manipulation, remote operation and research use. Avoid treating it as a modern commercial product.

Mobility: Bipedal humanoid research robot

Autonomy level: Research and supervised autonomy. It is not a commercial worker.

Why it matters: Valkyrie matters because NASA gave the US humanoid field a large research platform linked to Mars precursor work, disaster response and remote operation.

Limits and current status: Valkyrie is slow and research-oriented compared with current warehouse humanoids. It should be presented as a research platform with lessons for autonomy and operator interfaces.

Overview: Robonaut 2 is one of the most important American humanoid robots because it reached the International Space Station. It was built around dexterity and astronaut-adjacent work rather than walking.

Technical notes: The key hardware story is the upper body, hands and ability to use human tools. It should be classified separately from bipedal humanoids.

Mobility: Upper-body humanoid robot, originally torso-based for ISS work

Autonomy level: Telepresence and partial autonomy in research contexts.

Why it matters: Robonaut 2 matters because it proves that humanoid form can make sense when the environment is already designed for human hands, tools and procedures.

Limits and current status: Robonaut 2 is not a modern factory humanoid and should not be compared directly with Digit or Apollo on locomotion.

Overview: Nadia belongs in the US ecosystem because it connects IHMC’s long DRC-era humanoid control work with a modern research machine for whole-body tasks.

Technical notes: The page should describe Nadia through locomotion, manipulation, operator interfaces and behavior authoring instead of commercial deployment language.

Mobility: Bipedal humanoid research platform

Autonomy level: Research, supervised behavior and operator-directed tasks.

Why it matters: Nadia matters because many humanoid failures happen at the boundary between walking, reaching, contact and recovery. IHMC’s work is centered on that boundary.

Limits and current status: Nadia is a research robot. Do not list it as a consumer or factory product unless future sources confirm a commercial package.

Overview: THOR and THOR-OP represent the Virginia Tech side of the US humanoid research story. They belong with DARPA-era platforms, not modern commercial warehouse robots.

Technical notes: The robot profile should focus on whole-body disaster-response tasks, walking, manipulation and academic control research.

Mobility: Bipedal humanoid research platform

Autonomy level: Research and competition autonomy, often with operator involvement.

Why it matters: THOR matters because the DARPA Robotics Challenge forced teams to test humanoids on doors, valves, tools, rough terrain and degraded communications.

Limits and current status: Do not present THOR as an active commercial product. It is a research and competition-era platform.

Overview: CHARLI and CHARLI-2 show how American universities trained the generation that later built stronger humanoid systems. The platform is part of the academic layer of the ecosystem.

Technical notes: Keep the technical description focused on walking, balance, RoboCup and shipboard firefighting research connections.

Mobility: Bipedal humanoid research platform

Autonomy level: Academic research and demonstrations.

Why it matters: CHARLI matters because US humanoid robotics did not start with current startups. It came through university labs, competitions, Navy-funded research and years of bipedal control work.

Limits and current status: CHARLI is historical academic robotics, not a current commercial platform.