Causing a mobile robot to move according to a planned trajectory determined from a prediction of agent states of agents in an environment of the mobile robot

A method, comprising:

• producing, by a processor at a first time, a spatiotemporal graph representing relationships among agents in an environment of a mobile robot;
• inputting, by the processor, information associated with the spatiotemporal graph to neural networks to produce some information for a mixture of affine time-varying systems representing an evolution of agent states;
• calculating, by the processor using the mixture of affine time-varying systems and information associated with the first time, a prediction of the agent states at a second time;
• and
• causing, by the processor, the mobile robot to move according to a planned trajectory determined from the prediction.

A non-transitory computer-readable medium for causing a mobile robot to move according to a planned trajectory, the non-transitory computer-readable medium including instructions that when executed by one or more processors cause the one or more processors to:

• produce, at a first time, a spatiotemporal graph representing relationships among agents in an environment of the mobile robot;
• input information from the spatiotemporal graph to neural networks to produce some information for a mixture of affine time-varying systems representing an evolution of agent states;
• calculate, using the mixture of affine time-varying systems and information associated with the first time, a prediction of the agent states at a second time;
• and
• cause the mobile robot to move according to the planned trajectory determined from the prediction.