Configuration of robot operational environment including layout of sensors
A layout of a robotic operational environment is optimized to at least reduce or minimize occlusion of one or more sensors, to improve computational efficiency of motion planning, and/or improve an efficiency and/or robustness of generated motions plans. A respective sensor pose (i.e., positioned and oriented) for one or more sensors may be determined while taking into account the various robot poses of the robot(s) and transitions between robot poses. For instance, cost values that represent occlusion may be determined, and candidate sensor poses evaluated based on cost values or aggregated cost. The total number of sensors to be employed may not be known before execution, and in fact may be determined only after several iterations, upon reaching a termination condition (e.g., task completion time). Sensors make take any variety of forms, e.g., digital camera, radar, LIDAR, sonar, passive infrared (PIR) motion detector.
1 . A method of operation of a processor-based system to facilitate robot operation, the method comprising:
for each of a plurality of candidate sensor poses, performing an occlusion assessment that accounts for an occlusive effect of at least a first robot in an operational environment as the first robot moves through each of a plurality of robot poses of the first robot; and
selecting one of the plurality of the candidate sensor poses to pose a first sensor based at least in part on the assessment of occlusion,
wherein performing an occlusion assessment that accounts for an occlusive effect of at least a first robot in the operational environment as the first robot moves through each of a plurality of robot poses of the first robot includes, before a runtime and prior to performance of motion planning for the first robot to generate a motion plan, performing an occlusion assessment for each of a number of robot poses sampled from a roadmap of poses realizable by the first robot, the roadmap in a form of a graph comprising a set of nodes that represent respective robot poses in a C-space of the robot and a plurality of edges that couple pairs of the nodes and which represent transitions between the respective robot poses represented by the nodes which the respective edge couple, the method further comprising:
in an outer iterative loop executed for each voxel in a workspace,
determining a respective occlusion value for each of a number of other voxels in the workspace, where the occlusion values are representative of whether the voxel of a current iteration of the outer iterative loop is occluded from one or more sensors by an object positioned so as to at least partially occupy the respective voxel for which the occlusion value is being determined; and
aggregating a cost for the plurality of the candidate sensor poses.
2 . The method of claim 1 wherein the first sensor is a camera, the camera having a respective field-of-view of an operational environment for each of the plurality of the candidate sensor poses, and selecting one of the plurality of candidate sensor poses to pose the first sensor based at least in part on the assessment of occlusion includes selecting one of the plurality of the candidate sensor poses to pose the camera based at least in part on the assessment of occlusion.
3 . The method of claim 1 wherein performing an occlusion assessment that accounts for an occlusive effect of at least a first robot in the operational environment as the first robot moves through each of a plurality of robot poses of the first robot includes performing an occlusion assessment that accounts for an occlusive effect of the first robot and at least a second robot in the operational environment as the first and the second robots move through each of a respective plurality of robot poses of the first and the second robots.
4 . The method of claim 1 , further comprising:
for each of the plurality of candidate sensor poses, determining a respective cost value for the plurality of the candidate sensor pose, the respective cost value representative of occlusions.
5 . The method of claim 1 , further comprising:
before determining a respective occlusion value for the voxel,
for each candidate sensor pose, initializing a respective cost value of the candidate sensor pose;
in the outer iterative loop, for all voxels in the workspace, initializing the respective occlusion value of the voxel.
6 . The method of claim 1 wherein determining a respective occlusion value for the voxel of a current iteration of the outer iterative loop includes:
in an inner iterative loop, determining whether the voxel of the current iteration of the outer iterative loop is occluded based on any sensors already positioned;
in the inner iterative loop, if the voxel of the current iteration of the outer iterative loop is occluded, incrementing a respective occlusion value of any voxels occupied by an object in the form of a robot in one of the robot poses that is occluding the voxel of the current iteration of the outer iterative loop in the respective robot pose.
7 . The method of claim 1 wherein aggregating a cost for the plurality of the candidate sensor poses includes: for each candidate sensor pose, summing the occlusion values for each voxel between the sensor and the voxel of the current iteration of the outer iterative loop for which voxel occlusion is being assessed;
and adding the sum to an accumulated sum of costs for the respective candidate sensor pose.
8 . The method of claim 1 wherein selecting one of the plurality of the candidate sensor poses to pose a first sensor based at least in part on the assessment of occlusion includes selecting a candidate sensor pose having a lowest associated cost of the plurality of the candidate sensor poses relative to an associated cost of other ones of the candidate sensor poses.
9 . A method of operation of a processor-based system to facilitate robot operation, the method comprising:
for each of a plurality of candidate sensor poses, performing an occlusion assessment that accounts for an occlusive effect of at least a first robot in an operational environment as the first robot moves through each of a plurality of robot poses of the first robot; and
selecting one of the plurality of the candidate sensor poses to pose a first sensor based at least in part on the assessment of occlusion,
wherein performing an occlusion assessment that accounts for an occlusive effect of at least a first robot in the operational environment as the first robot moves through each of a plurality of robot poses of the first robot includes, before a runtime and prior to performance of motion planning for the first robot to generate a motion plan, performing an occlusion assessment for each of a number of robot poses sampled from a roadmap of poses realizable by the first robot, the roadmap in a form of a graph comprising a set of nodes that represent respective robot poses in a C-space of the robot and a plurality of edges that couple pairs of the nodes and which represent transitions between the respective robot poses represented by the nodes which the respective edge couple, the method further comprising:
after selecting one of the plurality of the candidate sensor poses to pose a first sensor based at least in part on the assessment of occlusion,
for each of the plurality of candidate sensor poses, performing an occlusion assessment that accounts for an occlusive effect of at least a first robot in the operational environment as at least the first robot moves through each of a plurality of robot poses of the first robot; and
selecting one of the plurality of the candidate sensor poses to pose a second sensor based at least in part on the assessment of occlusion.
10 . The method of claim 9 wherein performing an occlusion assessment that accounts for an occlusive effect of at least a first robot in the operational environment as the first robot moves through each of a plurality of robot poses of the first robot includes performing the occlusion assessment which also accounts for the first sensor positioned and oriented as specified by the selected one of the plurality of the candidate sensor poses for the first sensor.
11 . The method of claim 9 wherein the first and the second sensors are cameras, and further comprising:
receiving sensor information collected by a first one of the cameras in a first pose that corresponds to the selected sensor pose for the first sensor and from a second one of the cameras in a second pose that corresponds to the selected sensor pose for the second sensor;
performing motion planning for at least the first robot based on the received sensor information; and
driving the first robot to complete one of more task based on the motion planning.
12 . The method of claim 9 , further comprising:
after selecting one of the plurality of the candidate sensor poses to pose a second sensor based at least in part on the assessment of occlusion,
repeating until a termination condition is achieved, for each of a plurality of candidate sensor poses, performing an occlusion assessment that accounts for an occlusive effect of at least a first robot in the operational environment as at least the first robot moves through each of a plurality of robot poses of the first robot; and
selecting one of the plurality of the candidate sensor poses to pose an additional sensor based at least in part on the assessment of occlusion.
13 . A system of operation of a processor-based system to facilitate robot operation, the system comprising:
at least one processor;
at least one processor-readable medium communicatively coupled to the at least one processor and which stores processor-executable instructions that, when executed by the at least one processor, cause the at least one processor to:
for each of a plurality of candidate sensor poses, perform an occlusion assessment that accounts for an occlusive effect of at least a first robot in an operational environment as the first robot moves through each of a plurality of robot poses of the first robot; and
select one of the plurality of the candidate sensor poses to pose a first sensor based at least in part on the assessment of occlusion wherein, when executed, the processor-executable instructions cause the at least one processor further to:
in an outer iterative loop executed for each voxel in a workspace, determine a respective occlusion value for at least a number of other voxels in the workspace, where the occlusion values are representative of whether the voxel of a current iteration of the outer iterative loop is occluded from one or more sensors by an object positioned so as to at least partially occupy the respective voxel for which the occlusion value is being determined; and
aggregate a cost for the plurality of the candidate sensor poses.
14 . The system of claim 13 wherein to perform an occlusion assessment the instructions, when executed, cause the at least one processor to perform an occlusion assessment for each of a number of robot poses sampled from a roadmap of poses realizable by the first robot, the roadmap in a form of a graph comprising a set of nodes that represent respective robot poses in a C-space of the robot and a plurality of edges that couple pairs of the nodes and which represent transitions between the respective robot poses represented by the nodes which the respective edge couple.
15 . The system of claim 13 wherein to perform an occlusion assessment the instructions, when executed, cause the at least one processor to perform an occlusion assessment that accounts for an occlusive effect of the first robot and at least a second robot in the operational environment as the first and the second robots move through each of a respective plurality of robot poses of the first and the second robots.
16 . The system of claim 13 wherein, when executed, the processor-executable instructions cause the at least one processor further to:
for each of the plurality of the candidate sensor poses, determine a respective cost value for the candidate sensor pose, the respective cost value representative of occlusions.
17 . The system of claim 13 wherein, when executed, the processor-executable instructions cause the at least one processor further to:
before a respective occlusion value for the voxel is determined,
for each candidate sensor pose, initialize a respective cost value of the candidate sensor pose;
in the outer iterative loop, for all of the voxels in the workspace, initialize the respective occlusion value of the voxel.
18 . The system of claim 17 wherein to determine a respective occlusion value for the voxel of a current iteration of the outer iterative loop, when executed, the processor-executable instructions cause the at least one processor to:
in an inner iterative loop, determine whether the voxel of the current iteration of the outer iterative loop is occluded based on any sensors already positioned;
in the inner iterative loop, if the voxel of the current iteration of the outer iterative loop is occluded, increment a respective occlusion value of any voxels occupied by an object in a form of a robot in one of the robot poses that is occluding the voxel of the current iteration of the outer iterative loop in the respective robot pose.
19 . The system of claim 13 wherein to aggregate a cost for the plurality of the candidate sensor poses, when executed, the processor-executable instructions cause the at least one processor to:
for each candidate sensor pose, sum the occlusion values for each voxel between the sensor and the voxel of the current iteration of the outer iterative loop for which occlusion is being assessed in the current iteration of the outer iterative loop; and
add the sum to an accumulated sum of costs for the respective candidate sensor pose.
20 . The system of claim 13 wherein to select one of the candidate sensor poses to pose a first sensor based at least in part on the assessment of occlusion, when executed, the processor-executable instructions cause the at least one processor further to: select the candidate sensor pose having a lowest associated cost of the candidate sensor pose relative to an associated cost of other ones of the candidate sensor poses.
21 . The system of claim 13 wherein, when executed, the processor-executable instructions cause the at least one processor further to:
after selection of one of the plurality of the candidate sensor poses to pose a first sensor based at least in part on the assessment of occlusion,
for each of the plurality of candidate sensor poses, perform an occlusion assessment that accounts for an occlusive effect of at least a first robot in the operational environment as the first robot moves through each of a plurality of robot poses of the first robot; and select one of the plurality of the candidate sensor poses to pose a second sensor based at least in part on the assessment of occlusion.
22 . The system of claim 21 wherein to perform an occlusion assessment, when executed, the processor-executable instructions cause the at least one processor further to: perform the occlusion assessment which also accounts for the first sensor positioned and oriented as specified by the selected one of the candidate sensor poses for the first sensor.
23 . The system of claim 21 wherein the sensors are cameras, and wherein, when executed, the processor-executable instructions cause the at least one processor further to:
receive sensor information collected by a first one of the cameras in a first pose that corresponds to the selected sensor pose for the first sensor and from a second one of the cameras in a second pose that corresponds to the selected sensor pose for the second sensor; and
perform motion planning for at least the first robot based on the received sensor information.