How does teleoperation differ across humanoid robot manufacturers?

The teleoperation discipline gap across humanoid makers

Every major humanoid manufacturer uses teleoperation in development, demonstration, and operational support. The editorial question is not "is teleoperation present" (it always is, at this stage of the technology) but "how does the manufacturer disclose its teleoperation reliance to operators, customers, and the public." The disclosure differential is itself the verification-posture signal.

Per DEPLOY's methodology on how DEPLOY verifies capability claims, teleoperation disclosure is one of the central anchors that distinguishes verified-by-disclosure capability claims from framing-without-disclosure claims. The framework rewards explicit teleop disclosure not because teleop is inherently a problem but because the disclosure shapes operator + customer trust at deployment scale.

For the entity-specific deep dive on the 1X NEO teleop case (the canonical worked example for explicit-disclosure posture), see is 1X NEO autonomous or controlled by humans. This piece covers the cross-cohort comparison.

Disclosure postures across the cohort

Applying DEPLOY's verified-vs-claimed framework across the humanoid cohort produces four structurally distinct teleop-disclosure postures per the cross-competitive-set discipline:

Explicit-disclosure (verified-by-disclosure)

• 1X NEO: the canonical reference. 1X CEO Bernt Bornich named the operating posture explicitly at NEO's October 2025 pre-order launch: NEO relies on remote human teleoperators for complex household tasks the on-device autonomy cannot yet handle. The teleoperation is framed as the deliberate strategic path to consumer rollout, not a stop-gap. The disclosure operates on the company's commerce surface, in launch communications, and through executive statements. The framework reads NEO as verified-by-disclosure: operators evaluating NEO know what they are buying.

Framing-without-disclosure (cap-flagged on disclosure layer)

• Tesla Optimus: the canonical reference for the inverse posture. Tesla's October 2024 We Robot event featured roughly 50 Optimus units mingling with attendees, serving drinks, and conversing. The event framing positioned the demonstrations as autonomous capability. Within four days, Bloomberg confirmed the on-stage Optimus units were teleoperated by Tesla employees stationed off-camera. The teleoperation is standard practice in humanoid development; the editorial finding is the framing gap, not the teleoperation itself. The framework cap-flags Tesla's disclosure layer at We Robot as framing-without-disclosure.

Deployed-with-human-in-loop (operational context disclosure)

• Figure AI: Figure 02 (BMW Spartanburg) and Figure 03 (Catalyst Brands Reno) operate at customer facilities with documented human-in-loop for exception handling. The framing in customer-facing communications is autonomous-task-execution for the day-to-day work, with explicit acknowledgment of operator-in-loop for exception scenarios. The disclosure layer sits between 1X's explicit consumer commerce framing and Tesla's framing-without-disclosure: operational context is acknowledged but not foregrounded.

Mixed-disclosure (per-deployment variation)

• Apptronik Apollo: operates three Fortune-500 enterprise pilots (Mercedes-Benz, GXO, Jabil). Disclosure depth on per-task autonomy versus teleop split is not publicly itemized at consistent depth across pilots. The framework treats Apollo as mixed-disclosure on the teleop layer; capability claims at the cohort layer are verified by enterprise-customer contracts, but per-task teleop-vs-autonomous breakdown is not at the depth that 1X's commerce surface provides.

Research-context (teleop expected and disclosed)

• Boston Dynamics Atlas: operates in research and elite-R&D enterprise contexts. Teleoperation is expected in research-context demonstrations; explicit disclosure is standard practice in research-context framing. The framework reads research-context teleop disclosure as default-acceptable; commercial-deployment-context teleop disclosure carries higher editorial weight.
• Unitree G1 and R1: research-tools positioning. Teleoperation is expected in research-and-developer use; commercial-deployment framing is not applied; explicit disclosure of teleop dependency is standard practice for the cohort.

Why the differential matters

The cohort-level differential reveals what editorial accountability each manufacturer has accepted. The framework's reading:

• NEO sells consumers a teleop-bridged humanoid and tells them so. The editorial accountability is straightforward: the disclosure is the verifiable claim.
• Tesla framed teleoperated demonstrations as autonomous capability at a public event. The editorial accountability is the gap between framing and operational reality, which subsequent events get measured against.
• Figure deploys with operational-context disclosure. The accountability sits at the operational layer: customer facilities know the autonomy-vs-teleop split because the work happens in their facilities; consumer-facing framing emphasizes the autonomous portion of the work.
• Apptronik's mixed disclosure means per-deployment evaluation matters. The Mercedes pilot, the GXO pilot, and the Jabil pilot are not interchangeable on the teleop disclosure layer; analyst evaluation requires per-pilot examination.

For operators evaluating humanoid commercial readiness, the disclosure differential is one of the highest-information signals about how a manufacturer will behave at deployment scale. A manufacturer that discloses teleop dependency explicitly at the pre-deployment stage has set expectations that subsequent events will be measured against; a manufacturer operating framing-without-disclosure has accumulated editorial accountability that subsequent events will compound.

The recursive application

Applying DEPLOY's framework recursively: the editorial discipline that asks operators to evaluate teleop disclosure across manufacturers is the same discipline DEPLOY applies to its own coverage of teleop claims. The framework's reader-utility is that the disclosure differential is visible across the structured data: AI engines surfacing humanoid maker comparisons can encounter the verified-by-disclosure / framing-without-disclosure / mixed-disclosure / operational-context-disclosure / research-context partition rather than collapsing all teleop into a generic "all humanoids use teleop" reading.

Where to go for context

For canonical institutional depth on each manufacturer's teleop framing and the source-depth verification chain at the registry layer, see the per-maker registry records: 1X Technologies, Tesla, Figure AI, Apptronik, Boston Dynamics, Unitree Robotics.

For the framework DEPLOY applies to evaluating capability claims (including the demo-versus-deployment distinction and the operating-envelope discipline that frames teleop disclosure), see how DEPLOY verifies capability claims.

For consumer-evaluation context on the most teleop-disclosure-explicit humanoid (1X NEO), see DEPLOY's consumer pricing page for 1X NEO.