Abstract: Embodiments provide a method for obstacle avoidance in a mobile robot. Images of objects are captured with mobile robot image sensors with a floor level perspective. Data corresponding to the images is used to train a neural network. The identified objects are classified by indicating whether the objects are a potential hazard. In one embodiment, an image sensor on a robot provides intensity and depth data for each of a plurality of pixels of the images. The intensity and depth data are fused to produce fused data. The fused data is then used to train the neural network.

Abstract: Embodiments provide a method for obstacle avoidance in a mobile robot. Images of objects are captured with mobile robot image sensors with a floor level perspective. Data corresponding to the images is used to train a neural network. The identified objects are classified by indicating whether the objects are a potential hazard. In one embodiment, an image sensor on a robot provides intensity and depth data for each of a plurality of pixels of the images. The intensity and depth data are fused to produce fused data. The fused data is then used to train the neural network.

Abstract: In one embodiment, a user indicates one or more virtual zones on an area map on a user device for a particular robot. The zones are then transferred to the robot. The robot determines an optimum order for multiple zones and a path for navigating to the different zones.

Abstract: In one embodiment, a system and method is provided for optimizing the total time taken for cleaning an area by a cleaning robot, which is the sum of the cleaning time and the time taken for the robot to re-charge its battery mid-cleaning when the area is too large to clean on a single charge. In one embodiment, the remaining area to be cleaned is determined, and the mid-cleaning re-charge time is reduced to the amount of charge needed to clean the remaining area, plus a buffer. Hence, a reduction in mid-cleaning charging time results in reduced total time taken to clean a given space.

Abstract: In one embodiment, a cleaning robot stores multiple floorplans. The robot automatically determines which floorplan it is in by localizing and trying to match its detected environment to the stored map. The best fit is the map for the current floor. If there is no match above a confidence threshold, the robot assumes it is a new floor or the floorplan has changed (e.g., furniture has moved), and the robot initiates a discovery mode to map the floorplan and add it to the robot's memory.

Abstract: In one embodiment, a user indicates one or more virtual zones on an area map on a user device for a particular robot. The zones are then transferred to the robot. The robot determines an optimum order for multiple zones and a path for navigating to the different zones.

Abstract: In one embodiment, a cleaning robot stores multiple floorplans. The robot automatically determines which floorplan it is in by localizing and trying to match its detected environment to the stored map. The best fit is the map for the current floor. If there is no match above a confidence threshold, the robot assumes it is a new floor or the floorplan has changed (e.g., furniture has moved), and the robot initiates a discovery mode to map the floorplan and add it to the robot's memory.

Abstract: In one embodiment, a system and method is provided for optimizing the total time taken for cleaning an area by a cleaning robot, which is the sum of the cleaning time and the time taken for the robot to re-charge its battery mid-cleaning when the area is too large to clean on a single charge. In one embodiment, the remaining area to be cleaned is determined, and the mid-cleaning re-charge time is reduced to the amount of charge needed to clean the remaining area, plus a buffer. Hence, a reduction in mid-cleaning charging time results in reduced total time taken to clean a given space.

Abstract: An apparatus and method is presented for calibrating an output(s) of an inertial measurement unit (IMU) using rotational rate as a reference. Calibrating the IMU output(s) is performed by comparing the IMU output(s) to expected output(s), where the expected output(s) are determined based on the known rate of rotation of the IMU and the centripetal force acting on the IMU due to known rate of rotation. By analyzing the differences between the expected IMU output(s) and the IMU output(s), it is possible to determine a correction factor that, when applied to the IMU output(s), calibrates the IMU output(s) by correcting for measurement errors.

Abstract: An apparatus and method is presented for calibrating an output(s) of an inertial measurement unit (IMU) using rotational rate as a reference. Calibrating the IMU output(s) is performed by comparing the IMU output(s) to expected output(s), where the expected output(s) are determined based on the known rate of rotation of the IMU and the centripetal force acting on the IMU due to known rate of rotation. By analyzing the differences between the expected IMU output(s) and the IMU output(s), it is possible to determine a correction factor that, when applied to the IMU output(s), calibrates the IMU output(s) by correcting for measurement errors.