Abstract: A method is provided for maintaining an aircraft of a type of aircraft. The method includes accessing time series of observations of variables that describe in-service operation of the aircraft, maintenance of the aircraft, and weather and/or terrain in an environment of the aircraft during in-service operation and maintenance. The method includes determining a behavior model that is trained to predict demand for replacement aircraft parts from a training set across a plurality of aircraft of the type of aircraft. The behavior model is implemented as a directed acyclic graph of machine learning models that are connected to one another. The time series are applied to the behavior model to predict the demand for the replacement aircraft parts, a plan is created for acquisition of the replacement aircraft parts and maintenance of the aircraft based on the demand as predicted, and the plan is executed.

Abstract: A vacuuming device. The device includes a sensor connected to at least a suction head, and is configured to sense a parameter of foreign object debris (FOD) in the suction head. The device also includes a positioning sensor connected to the suction head and configured to ascertain a position of the suction head within an area in which the device is located. The device also includes a computer in communication with the sensor and the positioning sensor. The computer is programmed to classify the FOD using the parameter, and use the position of the suction head to determine a pick-up location of the FOD.

Abstract: A vacuuming device. The device includes a sensor connected to at least a suction head, and is configured to sense a parameter of foreign object debris (FOD) in the suction head. The device also includes a positioning sensor connected to the suction head and configured to ascertain a position of the suction head within an area in which the device is located. The device also includes a computer in communication with the sensor and the positioning sensor. The computer is programmed to classify the FOD using the parameter, and use the position of the suction head to determine a pick-up location of the FOD.

Abstract: A method is provided for maintaining an aircraft of a type of aircraft. The method includes accessing time series of observations of variables that describe in-service operation of the aircraft, maintenance of the aircraft, and weather and/or terrain in an environment of the aircraft during in-service operation and maintenance. The method includes determining a behavior model that is trained to predict demand for replacement aircraft parts from a training set across a plurality of aircraft of the type of aircraft. The behavior model is implemented as a directed acyclic graph of machine learning models that are connected to one another. The time series are applied to the behavior model to predict the demand for the replacement aircraft parts, a plan is created for acquisition of the replacement aircraft parts and maintenance of the aircraft based on the demand as predicted, and the plan is executed.