System and method for robotic object detection using a convolutional neural network
A system includes a mobile robot, the robot comprising a sensor; and a server operably connected to the robot over a network, the robot being configured to detect an object by processing sensor data using a convolutional neural network. A pipeline for robotic object detection using a convolutional neural network includes: a system comprising a mobile robot, the robot comprising a sensor, the system further comprising a server operably connected to the robot over a network, the robot being configured to detect an object by processing sensor data using a pipeline, the pipeline comprising a convolutional neural network, the pipeline configured to perform a data collection step, the pipeline further configured to perform a data transformation step, the pipeline further configured to perform a convolutional neural network step, the pipeline further configured to perform a network output transformation step, the pipeline further configured to perform a results output step.
1. A system comprising:
a robot having at least one sensor configured to provide data about an object, the robot configured to detect the object by processing sensor data using a convolutional neural network; and
a fleet management server including a processor connected to the robot and configured to train the convolutional neural network while the processor is not managing the robot when the processor is in an offline status,
wherein:
the convolutional neural network is configured to determine one or more of information about a location of the object and information about a type of the object, and
the type of the object is selected from the group consisting of: a vehicle, a robot, a cart, a landmark, a stationary object, a moving object, an autonomous guided vehicle, a forklift, a golf cart, and a warehouse.
2. The system of claim 1 wherein the at least one sensor is a 2D sensor.
3. The system of claim 2 wherein the at least one sensor is the 2D sensor and a 3D sensor.
4. The system of claim 1 wherein the data about the object includes one or more of a shape, a size, and a reflectivity of the object.
5. The system of claim 1 wherein the training comprises using a plurality of examples of an input to the convolutional neural network and a corresponding desired output from the convolutional neural network.
6. The system of claim 1 , wherein the at least one sensor comprises a light detection and ranging (LIDAR) sensor.
7. The system of claim 1 wherein the at least one sensor is further configured to detect an object that is onboard the robot.
8. A method comprising:
detecting an object, via a robot having at least one sensor, by processing sensor data using a convolutional neural network; and
training, via a fleet management server including a processor connected to the robot, the convolutional neural network while the processor is not managing the robot when the processor is in an offline status;
wherein:
the convolutional neural network is configured to determine one or more of information about a location of the object and information about a type of the object, and
the type of the object is selected from the group consisting of: a vehicle, a robot, a cart, a landmark, a stationary object, a moving object, an autonomous guided vehicle, a forklift, a golf cart, and a warehouse.
9. The method of claim 8 wherein the at least one sensor is a 2D sensor.
10. The method of claim 9 wherein the at least one sensor is the 2D sensor and a 3D sensor.
11. The method of claim 8 wherein the at least one sensor is a light detection and ranging (LIDAR) sensor.
12. The method of claim 8 wherein the training comprises using a plurality of examples of an input to the convolutional neural network and a corresponding desired output from the convolutional neural network.
13. The method of claim 8 further comprising detecting, via the at least one sensor, an object that is onboard the robot.
14. A non-transitory computer readable medium having stored thereon computer executable instructions, the instructions comprising:
detecting an object, via a robot having at least one sensor, by processing sensor data using a convolutional neural network; and
training, via a fleet management sever including a processor connected to the robot, the convolutional neural network while the processor is not managing the robot when the processor is in an offline status;
wherein:
the convolutional neural network is configured to determine one or more of information about a location of the object and information about a type of the object, and
the type of the object is selected from the group consisting of: a vehicle, a robot, a cart, a landmark, a stationary object, a moving object, an autonomous guided vehicle, a forklift, a golf cart, and a warehouse.
15. The non-transitory computer readable medium of claim 14 wherein the training comprises using a plurality of examples of an input to the convolutional neural network and a corresponding desired output from the convolutional neural network.
16. The non-transitory computer readable medium of claim 14 wherein the instructions further comprise detecting, via the at least one sensor, an object that is onboard the robot.
17. The system of claim 1 wherein the type of the object further comprises a shelf.
18. The method of claim 8 wherein the type of the object further comprises a shelf.