Hardware & control
Motion planning
Motion planning is the computation of a feasible path or trajectory that moves a robot from an initial condition toward a goal while satisfying constraints such as collision avoidance, joint limits, contact, balance, or dynamics. The planner searches over possible robot motions; a controller is still needed to execute the result.
Also known as: robot motion planning, robot path planning
Updated
Planning searches for a feasible motion
LaValle's Planning Algorithms treats motion planning through configuration spaces, constraints and search. A planner may work in joint space, task space or a combined state space. Sampling-based, search-based and optimisation-based methods explore that space in different ways.
A geometric path specifies an ordered route without timing. A trajectory adds time and may include velocity or acceleration. Dynamic planning must account for how controls and physical dynamics constrain the reachable motion.
A collision-free arm path may still fail a humanoid
Humanoid motion can require balance, contact schedules, whole-body coordination, self-collision avoidance and actuator limits. Planning only the hand path does not ensure that the body can follow it while keeping its feet stable.
The environment model also matters. Missing geometry, pose uncertainty or a moving person can invalidate a plan. Safety margins help but do not replace sensing and replanning.
Planning and control are separate layers
A planner proposes a path or trajectory; feedback control tracks it and responds to deviations. Model predictive control can repeatedly optimise a short horizon and therefore blur the boundary, but the execution loop still depends on state estimation and low-level control.
Planning datasets should retain start and goal conditions, robot and environment models, constraints, collision geometry, planner settings, costs, random seeds, success status and computation time. A smooth successful trajectory alone hides failed searches and says little about robustness.
Sources
Related terms
Data & collection
Trajectory
A trajectory is a time-ordered sequence of states or observations, actions and, where applicable, rewards generated as an agent or robot evolves. A complete episode or policy rollout often yields a trajectory, but the terms are not universally identical: trajectories may be partial, while episodes have dataset- or environment-defined boundaries.
Hardware & control
Joint space
Joint space is a coordinate space whose variables describe a robot's joint configuration, such as revolute-joint angles and prismatic-joint displacements. A point represents one configuration subject to the robot model and joint limits; a path or trajectory represents how that configuration changes.
Hardware & control
Task space
Task space is a coordinate space used to express quantities directly relevant to a robot task, such as an end-effector pose, centre-of-mass position, gaze direction, contact force, or several objectives together. It describes what should be achieved, while joint space describes the robot configuration used to achieve it.
Simulation & transfer
Collision detection
Collision detection determines whether geometric representations of a robot, itself, or its environment intersect or come within a specified distance. Robotics systems use it for self-collision checking, environment collision checking, motion planning, simulation, and safety monitoring, but its result depends on the geometry, poses, margins, and update timing supplied.
Hardware & control
Robot kinematics
Robot kinematics describes the geometric relationship between a robot’s joint configuration and the position, orientation and velocity of its links or end-effector, without modelling the forces that cause the motion. Forward kinematics computes pose from joint values; inverse kinematics searches for joint values that achieve a requested pose.
Hardware & control
Model predictive control
Model predictive control is a feedback-control method that uses a dynamics model to optimise a sequence of future actions, applies the first action or short part of that sequence, and replans from an updated state estimate. In receding-horizon MPC, the planning window moves forward at each update.