Topology processing for waypoint-based navigation maps
The operations of a computer-implemented method include obtaining a topological map of an environment including a series of waypoints and a series of edges. Each edge topologically connects a corresponding pair of adjacent waypoints. The edges represent traversable routes for a robot. The operations include determining, using the topological map and sensor data captured by the robot, one or more candidate alternate edges. Each candidate alternate edge potentially connects a corresponding pair of waypoints that are not connected by one of the edges. For each respective candidate alternate edge, the operations include determining, using the sensor data, whether the robot can traverse the respective candidate alternate edge without colliding with an obstacle and, when the robot can traverse the respective candidate alternate edge, confirming the respective candidate alternate edge as a respective alternate edge. The operations include updating, using nonlinear optimization and the confirmed alternate edges, the topological map.
1 . A computer-implemented method comprising:
obtaining, by data processing hardware of a robot, a topological map of an environment, wherein the topological map indicates a first route waypoint, a second route waypoint, a third route waypoint, a fourth route waypoint, a first route edge, and a second route edge collectively used by the robot to indicate a route, wherein the first route edge directly connects the first route waypoint and the second route waypoint, and wherein the second route edge directly connects the second route waypoint and the third route waypoint;
determining, by the data processing hardware, using the topological map and first sensor data captured by the robot, a third route edge, wherein the third route edge directly connects the second route waypoint and the fourth route waypoint;
determining, by the data processing hardware, using the first sensor data, that the third route edge is traversable by the robot;
updating, by the data processing hardware, using the third route edge, the topological map to obtain an updated topological map based on determining that the third route edge is traversable by the robot; and
instructing, by the data processing hardware, navigation of the robot according to the updated topological map.
2 . The method of claim 1 , further comprising:
generating data indicating at least one of the first route waypoint, the second route waypoint, the third route waypoint, the fourth route waypoint, the first route edge, or the second route edge using odometry data captured by the robot.
3 . The method of claim 1 , wherein at least one of the first route waypoint, the second route waypoint, the third route waypoint, or the fourth route waypoint is associated with respective second sensor data captured by the robot.
4 . The method of claim 1 , wherein determining the third route edge comprises:
determining, using the topological map, a local embedding;
determining that a total path length between the second route waypoint and the fourth route waypoint satisfies a first threshold distance;
determining that a distance in-associated with the local embedding satisfies a second threshold distance; and
generating the third route edge based on determining that the total path length between the second route waypoint and the fourth route waypoint satisfies the first threshold distance and determining that the distance associated with the local embedding satisfies the second threshold distance.
5 . The method of claim 1 , wherein determining that the third route edge is traversable by the robot comprises:
determining that the third route edge is traversable by the robot further using an output of a sensor data alignment algorithm.
6 . The method of claim 5 , wherein the sensor data alignment algorithm comprises at least one of an iterative closest point algorithm, a feature-matching algorithm, a normal distribution transform algorithm, a dense image alignment algorithm, or a primitive alignment algorithm.
7 . The method of claim 1 , wherein determining that the third route edge is traversable by the robot comprises:
determining, using the topological map, a local embedding; and
determining that the third route edge is traversable by the robot based on the local embedding and a path collision checking algorithm.
8 . The method of claim 7 , wherein the path collision checking algorithm comprises a circle sweep algorithm.
9 . The method of claim 1 , wherein determining the third route edge comprises:
determining an embedding using a fiducial marker; and
determining the third route edge based on the embedding.
10 . The method of claim 1 , wherein updating the topological map comprises:
correlating a particular route waypoint with a metric location.
11 . The method of claim 1 , wherein updating the topological map comprises:
determining an embedding using sparse nonlinear optimization.
12 . The method of claim 1 , wherein updating the updated topological map that is metrically consistent.
13 . A system comprising:
data processing hardware; and
memory hardware in communication with the data processing hardware, the memory hardware storing instructions, wherein execution of the instructions by the data processing hardware causes the data processing hardware to:
obtain a topological map of an environment, wherein the topological map indicates a first route waypoint, a second route waypoint, a third route waypoint, a fourth route waypoint, a first route edge, and a second route edge collectively used by a robot to indicate a route, wherein the first route edge directly connects the first route waypoint and the second route waypoint, and wherein the second route edge directly connects the second route waypoint and the third route waypoint;
determine, using the topological map and first sensor data captured by the robot, a third route edge, wherein the third route edge directly connects the second route waypoint and the fourth route waypoint;
determine, using the first sensor data, that the third route edge is traversable by the robot;
update, using the third route edge, the topological map to obtain an updated topological map based on determining that the third route edge is traversable by the robot; and
instruct navigation of the robot according to the updated topological map.
14 . The system of claim 13 , wherein the execution of the instructions by the data processing hardware further causes the data processing hardware to:
generate data indicating at least one of the first route waypoint, the second route waypoint, the third route waypoint, the fourth route waypoint, the first route edge, or the second route edge using odometry data captured by the robot.
15 . The system of claim 13 , wherein at least one of the first route waypoint, the second route waypoint, the third route waypoint, or the fourth route waypoint is associated with a respective second sensor data captured by the robot.
16 . The system of claim 13 , wherein to determine the third route edge, the execution of the instructions by the data processing hardware further causes the data processing hardware to:
determine, using the topological map, a local embedding;
determine that a total path length between the second route waypoint and the fourth route waypoint satisfies a first threshold distance;
determine that a distance associated with the local embedding satisfies a second threshold distance; and
generate the third route edge based on determining that the total path length between the second route waypoint and the fourth route waypoint satisfies the first threshold distance and determining that the distance associated with the local embedding satisfies the second threshold distance.
17 . The system of claim 13 , wherein to determine that the third route edge is traversable by the robot, the execution of the instructions by the data processing hardware further causes the data processing hardware to:
determine that the third route edge is traversable by the robot further using an output of a sensor data alignment algorithm.
18 . The system of claim 17 , wherein the sensor data alignment algorithm comprises at least one of an iterative closest point algorithm, a feature-matching algorithm, a normal distribution transform algorithm, a dense image alignment algorithm, or a primitive alignment algorithm.
19 . The system of claim 13 , wherein to determine that the third route edge is traversable by the robot, the execution of the instructions by the data processing hardware further causes the data processing hardware to:
determine, using the topological map, a local embedding; and
determine that the third route edge is traversable by the robot based on the local embedding and a path collision checking algorithm.
20 . The system of claim 19 , wherein the path collision checking algorithm comprises a circle sweep algorithm.
21 . The system of claim 13 , wherein to determine the third route edge, the execution of the instructions by the data processing hardware further causes the data processing hardware to:
determine an embedding using a fiducial marker; and
determine the third route edge based on the embedding.
22 . The system of claim 13 , wherein to update the topological map, the execution of the instructions by the data processing hardware further causes the data processing hardware to:
correlate a particular route waypoint with a metric location.
23 . The system of claim 13 , wherein to update the topological map, the execution of the instructions by the data processing hardware further causes the data processing hardware to:
determine an embedding using sparse nonlinear optimization.
24 . The system of claim 13 , wherein the updated topological map is metrically consistent.