Humanoid introduces KinetIQ AI framework for humanoid robot orchestration

UK-based AI and robotics company Humanoid launched KinetIQ, which the company describes as an AI framework for end-to-end orchestration of heterogeneous humanoid robot fleets across industrial, service and home applications.

Humanoid said that KinetIQ acts as a single system that controls robots with different embodiments and coordinates the interactions between them.

KinetIQ features and architecture overview

The company said that the architecture is cross-timescale: four layers operate simultaneously, from fleet-level goal assignment down to millisecond-level joint control. Each layer treats the layer below as a set of tools, orchestrating them via prompting and tool use to achieve goals set from above. This agentic pattern, which Humanoid said is proven in frontier AI systems, lets components improve independently while the overall system scales naturally to larger fleets and more complex tasks.

Humanoid said its wheeled-base robots run industrial workflows: back-of-store grocery picking, container handling, and packing across retail, logistics and manufacturing. The company’s bipedal robot is its R&D platform for service and home, showcasing voice interaction, online ordering and grocery handling as an intelligent assistant.

The company highlighted the following two architectural elements of KinetIQ:

• Cross-embodiment
  • A single AI model can control robots with different morphologies and end-effector designs. Data collected on one embodiment helps train and improve performance across the fleet.
• Cross-timescale
  • KinetIQ simultaneously operates across several cognitive layers, each running at its own timescale, both in terms of the decision-making frequency and planning horizon.

System elements of KinetIQ

Humanoid provided the following details of the various system-level management of KinetIQ.

System 3 - Humanoid AI fleet agent

An agentic AI layer that treats each robot as a tool in its repertoire and reacts within seconds to use them and optimize fleet operations.

Humanoid said that System 3 integrates with facility management systems across logistics, retail and manufacturing, and is also applicable to service scenarios and smart-home coordination. The KinetIQ Agentic Fleet Orchestrator ingests task requests, expected outcomes, SOPs, real-time request updates and facility context, and allocates tasks and information across wheeled and bipedal robots, coordinating robot swaps at workstations, to maximize throughput and uptime.

The KinetIQ Fleet Orchestrator directs two-way communication with facility systems to:

• Receive new task requests and changes/reassignments
• Track task progress and performance metrics
• Report completion and issues
• Ensure exceptions are handled and resolved in coordination with traditional or agentic facility management systems.

System 2 - Robot-level reasoning

Humanoid describes System 2 as a robot-level agentic layer that plans interactions with the environment to achieve goals set by System 3. It spans a second to sub-minute timescale.

System 2 uses an omni-modal language model to observe the environment and interpret high-level instructions from System 3. It decomposes goals into sub-tasks by reasoning about the required actions to complete its assignments, as well as the best sequence and approach.

The company said that plans are updated dynamically from visual context rather than by relying on fixed, pre-programmed sequences, similar to how agentic systems select and sequence tools. These plans can be saved as workflows/SOPs to be executed again in the future and shared across the fleet.

System 2 also monitors execution and evaluates whether the VLA (System 1) is making progress. If the system determines it is unable to complete a task or needs assistance, it requests human support through the fleet layer (System 3). Humanoid said that assistance can be delivered via interventions at the System 2 level (through prompting) or at the System 1 level (through teleoperation or direct joint control), either remotely or on-site.

System 1 - VLA-based low-level task execution

A Vision-Language-Action (VLA) neural network that commands target poses for a subset of robot body parts (e.g., hands, torso or pelvis), driving progress toward immediate low-level objectives set by System 2.

System 1 exposes multiple low-level capabilities to System 2 that can be invoked via different prompts. Examples include picking & placing objects, manipulating containers, packing or locomoting. VLM-based reasoning of System 2 selects the capability most appropriate for the current situation and the goal. Each low-level capability is also capable of reporting its status (success, failure or in-progress) back to System 2 to facilitate progress tracking.

Humanoid said that KinetIQ VLA emits new predictions at a sub-second timescale, usually 5-10Hz. Each prediction constitutes a chunk of higher-frequency actions (30 to 50Hz, depending on the task) that will be executed by System 0. Action execution is fully asynchronous: a new action chunk is always being prepared while the previous one is still executed. To ensure that an asynchronously produced chunk doesn’t contradict the reality that unfolded while it was produced, the company said that KinetIQ uses the prefix conditioning technique: every chunk prediction is conditioned on the part of the previous chunk that is expected to be executed during inference.

Unlike impainting, Humanoid said that this is a universal technique equally applicable to both autoregressive and flow-matching models.

System 0 - RL-based whole-body control

Humanoid said that the goal of System 0 is to achieve pose targets set by System 1, while solving for the state of all robot joints in a way that continuously guarantees dynamic stability. System 0 runs at 50 Hz.

KinetIQ implementation of System 0 uses reinforcement learning (RL)-trained whole-body control for both bipedal and wheeled robots. The company said that this approach allows KinetIQ to fully exploit the synergy between different platforms, benefitting from the power of RL in producing capable locomotion controllers.

Whole body control is trained solely in simulation with online reinforcement learning, requiring roughly 15k hours of experience to produce a capable model.

Humanoid said that by working in unison across multiple embodiments and timescales, the four cognitive layers of KinetIQ can achieve complex goals that require fleet orchestration, reasoning, dexterous manipulation, dynamic recovery and stability control.

The company said that KinetIQ’s fully-agentic design, which embraces recent breakthroughs in the field of AI is one of the key factors behind Humanoid’s rapid progress towards solving Physical AI.