Models & learning
Multi-robot collaboration
Multi-robot collaboration is coordinated work by two or more robots toward a shared task outcome, where one robot's actions, timing, or state can change what another robot can observe or do.
Also known as: multi-robot cooperation, robot-robot collaboration
Updated
Multi-robot collaboration
Multi-robot collaboration occurs when two or more robots coordinate their actions toward a shared task outcome. Each robot can have a different role, and one robot's timing or state can change what another can observe or do.
A shared task is different from parallel work
Two robots moving unrelated objects at the same time are working in parallel. A handoff, a shared carry, or one robot holding a door while another moves a load is collaborative because the task depends on both agents' timing and the state they change together. The MATE preprint studies this pattern with synchronized two-humanoid episodes in a shared simulation.
Keep both agents in the data
Training records for collaboration should preserve both robots' observations and actions on a common timeline, along with their roles and shared object or environment state. Treating each trajectory as independent can remove the partner action that made a handoff or cooperative movement possible.
The term does not specify one communication protocol or control architecture. A partner may be teleoperated, scripted, or controlled by another policy; a dataset should identify which is true during collection and evaluation. “Multi-robot collaboration” describes the relationship between the agents, not a guarantee that they were autonomous or physically deployed together.
Sources
Related terms
Hardware & control
Bimanual manipulation
Bimanual manipulation is the use of two hands or robot manipulators to carry out one manipulation task with spatial, temporal, or force coordination between them. The two sides may play symmetric roles, such as lifting a large object together, or asymmetric roles, such as one hand holding an object while the other operates a tool.
Data & collection
Demonstration
A demonstration is a recorded example of how an intended task or behaviour is performed, usually represented as a time-aligned sequence of observations, states and actions. For humanoid robot learning, demonstrations may come from teleoperation, kinaesthetic guidance, motion capture or autonomous experts and provide targets for imitation.
Data & collection
Human–robot interaction
Human–robot interaction is the exchange of information and actions between people and robots while performing a task, and the field that studies how such systems are designed and evaluated. It includes remote operation, supervision, collaboration, physical interaction and social communication rather than only face-to-face conversation.
Models & learning
Long-horizon task
A long-horizon task is a temporally extended robot task whose success depends on maintaining reliable behaviour across many actions, phases, or dependent subtasks. The term has no universal step-count threshold: it usually signals sequential dependencies, accumulating execution error, delayed outcomes, changing object state, or information that must be remembered beyond the current observation.
Data & collection
Teleoperation
Teleoperation is real-time human control of a robot from a remote or mediated interface. For humanoid training-data collection, the operator’s inputs are mapped to robot motions while cameras, proprioception, commands and outcomes are recorded, producing embodied demonstrations in the robot’s own observation and action spaces.