Dynamic lane biasing

A method comprising:

• receiving sensor data from a sensor associated with a vehicle traversing an environment;
• determining a first area for the vehicle to traverse in the environment, the first area associated with a first width;
• determining an object represented in the sensor data;
• determining a distance between the vehicle and the object;
• determining a classification associated with the object;
• determining, based at least in part on the classification, a width of a lane bias region of a plurality of widths, wherein the width is based at least in part on a classification;
• determining, based at least in part on the classification, a second area associated with a second width that is less than the first width by the width of the lane bias region;
• determining, based at least in part on the second area, a cost associated with the vehicle traversing the second area;
• determining, based at least in part on the cost, a trajectory;
• and
• controlling the vehicle based at least in part on the trajectory.

A non-transitory computer-readable medium storing instructions executable by one or more processors, wherein the instructions, when executed, cause the one or more processors to perform operations comprising:

• receiving sensor data from a sensor associated with a vehicle traversing an environment;
• determining a first area on which the vehicle is able to traverse, the first area associated with a first width;
• determining an object represented in the sensor data;
• determining a classification associated with the object;
• determining, based at least in part on the classification, a second area associated with a second width;
• determining, based at least in part on the second area, a cost associated with operating the vehicle in the environment;
• determining, based at least in part on the cost, a trajectory associated with the vehicle;
• and
• controlling the vehicle based at least in part on the trajectory.