Service module for warehouse traffic planning assistance
One example method includes receiving, by an integrating hub from each source in a group of sources, a signal and position information, so that multiple signals and respective position information for the signals is received, generating, by the integrating hub, a full risk map, and the full risk map is generated using the signals and their respective position information, and by the integrating hub, making the full risk map available to the sources, wherein the full risk map is usable by the sources to guide operations of the sources in an operating environment.
1 . A method comprising:
receiving, by an integrating hub and from each source in a group of sources, a corresponding signal and corresponding position information, so that multiple signals and respective position information for the signals are received, wherein at least some of the multiple signals and position information are received in a form of a light detection and ranging (LiDAR) point cloud;
transforming the LiDAR point cloud into a three-dimensional (3D) map model that yields a hierarchical discrete representation of a space using voxels, wherein said transforming includes:
generating, from LiDAR data points obtained from the LiDAR point cloud, an octomap;
determining an aisle width within the space by performing ray-casting laterally from an origin point on the octomap and sweeping vertically and horizontally to compute a Euclidean distance projected in an x-y plane; and
classifying the aisle width using a threshold to produce a boolean narrowness flag for the position;
generating, by the integrating hub, a full risk map, wherein the full risk map is generated using the signals and their respective position information and is generated using the 3D map model, which represents the space using the voxels, wherein the boolean narrowness flag is incorporated into the full risk map as a binary indication for a quadrant position, wherein the full risk map is also generated by (i) identifying two signals that are of a same type and that share common position information and (ii) averaging the two signals to create an average signal, and wherein the average signal is used in the full risk map and is used to represent signal information for the common position information; and
making, by the integrating hub, the full risk map available to the sources, wherein the sources use the full risk map to guide operations in an operating environment, which includes said space.
2 . The method as recited in claim 1 , wherein each of the signals was generated by a respective sensor, and the position information for each of the signals indicates where, in the operating environment, the corresponding source was located when the signal was generated.
3 . The method as recited in claim 1 , wherein each of the signals was generated based on the presence of a corresponding feature of the operating environment.
4 . The method as recited in claim 1 , wherein each of the signals was generated based on either a transitory physical feature of the operating environment, or a permanent physical feature of the operating environment.
5 . The method as recited in claim 1 , wherein one of the sources comprises an autonomous mobile robot, or a non-autonomous vehicle, with sensors and position information acquisition capability.
6 . The method as recited in claim 1 , wherein the signal and the position information received by the integrating hub from one of the sources is received by the integrating hub in a form of a partial risk map.
7 . The method as recited in claim 1 , wherein the signal and the position information received by the integrating hub from one of the sources is received by the integrating hub in a form of a dictionary.
8 . The method as recited in claim 1 , wherein the full risk map is generated by the integrating hub using a partial risk map generated by the integrating hub.
9 . The method as recited in claim 1 , wherein one of the signals was generated by a range-based sensor.
10 . The method as recited in claim 1 , wherein the full risk map is a two dimensional map.
11 . A non-transitory storage medium having stored therein instructions that are executable by one or more hardware processors to perform operations comprising:
receiving, by an integrating hub and from each source in a group of sources, a corresponding signal and corresponding position information, so that multiple signals and respective position information for the signals are received, wherein at least some of the multiple signals and position information are received in a form of a light detection and ranging (LiDAR) point cloud;
transforming the LiDAR point cloud into a three-dimensional (3D) map model that yields a hierarchical discrete representation of a space using voxels, wherein said transforming includes:
generating, from LiDAR data points obtained from the LiDAR point cloud, an octomap;
determining an aisle width within the space by performing ray-casting laterally from an origin point on the octomap and sweeping vertically and horizontally to compute a Euclidean distance projected in an x-y plane; and
classifying the aisle width using a threshold to produce a boolean narrowness flag for the position;
generating, by the integrating hub, a full risk map, wherein the full risk map is generated using the signals and their respective position information and is generated using the 3D map model, which represents the space using the voxels, wherein the boolean narrowness flag is incorporated into the full risk map as a binary indication for a quadrant position, wherein the full risk map is also generated by (i) identifying two signals that are of a same type and that share common position information and (ii) averaging the two signals to create an average signal, and wherein the average signal is used in the full risk map and is used to represent signal information for the common position information; and
making, by the integrating hub, the full risk map available to the sources, wherein the sources use the full risk map to guide operations in an operating environment, which includes said space.
12 . The non-transitory storage medium as recited in claim 11 , wherein each of the signals was generated by a respective sensor, and the position information for each of the signals indicates where, in the operating environment, the corresponding source was located when the signal was generated.
13 . The non-transitory storage medium as recited in claim 11 , wherein each of the signals was generated based on the presence of a corresponding feature of the operating environment.
14 . The non-transitory storage medium as recited in claim 11 , wherein each of the signals was generated based on either a transitory physical feature of the operating environment, or a permanent physical feature of the operating environment.
15 . The non-transitory storage medium as recited in claim 11 , wherein one of the sources comprises an autonomous mobile robot, or a non-autonomous vehicle, with sensors and position information acquisition capability.
16 . The non-transitory storage medium as recited in claim 11 , wherein the signal and the position information received by the integrating hub from one of the sources is received by the integrating hub in a form of a partial risk map.
17 . The non-transitory storage medium as recited in claim 11 , wherein the signal and the position information received by the integrating hub from one of the sources is received by the integrating hub in a form of a dictionary.
18 . The non-transitory storage medium as recited in claim 11 , wherein the full risk map is generated by the integrating hub using a partial risk map generated by the integrating hub.
19 . The non-transitory storage medium as recited in claim 11 , wherein one of the signals was generated by a range-based sensor.
20 . The method of claim 1 , wherein a subset of the multiple signals and position information represents a sloped region of the operating environment, and wherein the method further includes:
defining a slope threshold that dictates whether a given slope of a given sloped region is classified as a hazard;
applying the slope threshold to the sloped region of the operating environment; and
modifying the full risk map by inserting the sloped region of the operating environment into the full risk map.