Motion prediction

A computer system comprising:

• at least one memory embodying executable instructions;
• at least one hardware processor coupled to the at least one memory and configured to predict behaviour of an agent for executing an objective of a mobile robot in a vicinity of the agent in dependence on the predicted behaviour, by:
• determining a reference path, wherein multiple actions are available to the agent, and the reference path relates to one of those actions;
• projecting a measured velocity vector of the agent onto the reference path, thereby determining a projected speed value for the agent along the reference path;
• computing predicted agent motion data for the agent along the reference path based on the projected speed value;
• assessing, by a planner, a plurality of possible ego trajectories for the mobile robot for fulfilling the objective of the mobile robot, using ego motion data of each possible ego trajectory and the predicted agent motion data along the reference path;
• generating a series of control signals for controlling a mobile robot to fulfil the objective in dependence on the predicted agent motion data;
• and
• controlling motion of the mobile robot in response to the series of control signals,
• wherein a wait period commences upon the planner first determining that there exists an acceptable trajectory for fulfilling the objective, based on the predicted agent motion, but the planner does not attempt to fulfil the objective until the wait period has expired, with at least one further iteration of the projecting and the computing being performed using an updated measured velocity of the agent for determining whether an acceptable trajectory for fulfilling the objective still exists upon expiry of the wait period, the control signals for fulfilling the objective only being generated once the wait period has expired and only if an acceptable trajectory still exists,
• wherein a controller generates at least an initial control signal of the series of control signals in response to the planner outputting, to the controller, the acceptable trajectory at an end of the wait period, but the planner continues to assess possible ego trajectories and subsequent control signal(s) of the series of control signals are generated in dependence on that continued assessment;
• wherein the wait period only commences if it is determined that the agent satisfies at least one measurement certainty condition, and
• wherein if the at least one measurement certainty condition is satisfied at the commencement of the wait period, but stops being satisfied during the wait period, the wait period resets and does not commence again until the measurement certainty condition is once again satisfied.

Non-transitory media embodying executable instructions configured, upon execution by at least one hardware processor, to perform a method for predicting behaviour of an agent for executing an objective of a mobile robot in a vicinity of the agent in dependence on the predicted behaviour, the method comprising:

• determining a reference path, wherein multiple actions are available to the agent, and the reference path relates to one of those actions;
• projecting a measured velocity vector of the agent onto the reference path, thereby determining a projected speed value for the agent along the reference path;
• computing predicted agent motion data for the agent along the reference path based on the projected speed value;
• assessing, by a planner, a plurality of possible ego trajectories for the mobile robot for fulfilling the objective of the mobile robot, using ego motion data of each possible ego trajectory and the predicted agent motion data along the reference path;
• generating a series of control signals for controlling a mobile robot to fulfil the objective in dependence on the predicted agent motion data;
• and
• controlling motion of the mobile robot in response to the series of control signals,
• wherein a wait period commences upon the planner first determining that there exists an acceptable trajectory for fulfilling the objective, based on the predicted agent motion, but the planner does not attempt to fulfil the objective until the wait period has expired, with at least one further iteration of the projecting and the computing being performed using an updated measured velocity of the agent for determining whether an acceptable trajectory for fulfilling the objective still exists upon expiry of the wait period, the control signals for fulfilling the objective only being generated once the wait period has expired and only if an acceptable trajectory still exists,
• wherein a controller generates at least an initial control signal of the series of control signals in response to the planner outputting, to the controller, the acceptable trajectory at an end of the wait period, but the planner continues to assess possible ego trajectories and subsequent control signal(s) of the series of control signals are generated in dependence on that continued assessment,
• wherein the wait period only commences if it is determined that the agent satisfies at least one measurement certainty condition, and
• wherein if the at least one measurement certainty condition is satisfied at the commencement of the wait period, but stops being satisfied during the wait period, the wait period resets and does not commence again until the measurement certainty condition is once again satisfied.