Abstract: Described is a system for robot supervisory control. The system includes an operator device that receives camera imagery from a camera mounted on a robot and three-dimensional (3D) data from a 3D sensor mounted on the robot. The user interface displays a two-dimensional (2D) view of the scene from the camera and 3D sensor data. One or more object markers of objects in the scene are overlaid on the 2D view or 3D sensor data. Viewing the scene, the user can choose a robotic action from a library of actions, which generates a simulation of the robot performing the selected action. The simulation is then rendered and overlaid on top of the 3D sensor data. Moreover, the simulation shows the expected variability of the robot's action. Based on the simulation's outcome, the user can approve and trigger the execution of the action in the real robot.

Abstract: What is illustrated and described is a pump assembly (1) for a pump., particularly for a miniature pump, with an electric motor (2) and with a housing unit (3), wherein the electric motor (2) has a motor shaft (5) passing freely through a motor housing of the electric motor (2) and the motor housing is connected to the housing unit (3), and wherein the motor shaft (5) is supported in the motor housing with an axial shaft play. According to the invention, at least one stabilizing element (6, 13, 14, 16) arranged outside of the motor housing is provided in order to reduce or eliminate the axial play of the motor shaft (5), with an axial compressive force or axial tensile force acting on the motor shaft (3) being applied by the stabilizing element (6, 13, 14, 16) to the motor shaft (5).

Abstract: A method for merging 3D point clouds from sparsely distributed viewpoints includes collecting a plurality of 3D point cloud data sets using a 3D sensor, each 3D point cloud data set in a local reference frame of a viewpoint of the 3D sensor, downsampling the 3D point cloud data sets, registering the downsampled 3D point cloud data sets to a global reference frame using an initial transform Tv for rotating and translating each downsampled 3D point cloud data set from the local reference frame to the global reference frame, deriving estimated transforms Tv until the last derived transform Tv converges to a stable transform Tv, registering the plurality of 3D point cloud data sets to the global reference frame using the stable transform Tv, and deriving second estimated transforms Tv until the last derived transform Tv converges to a second stable transform Tv.

Abstract: Described is a system for detecting attacks on mobile networks. The system includes the relevant hardware and components to perform a variety of operations including continuously measuring time-varying signals at each node in a network. The system determines network flux on the time-varying signals of all nodes in the network and detects a network attack if the network flux exceeds a predetermined threshold. Further, a reactive protocol is initiated if the network flux exceeds the predetermined threshold.

Abstract: Described is a system for system for gait intervention and fall prevention. The system is incorporated into a body suit having a plurality of distributed sensors and a vestibulo-muscular biostim array. The sensors are operable for providing biosensor data to the analytics module, while the vestibulo-muscular biostim array includes a plurality of distributed effectors. The analytics module is connected with the body suit and sensors and is operable for receiving biosensor data and analyzing a particular user's gait and predicting falls. Finally, a closed-loop biostim control module is included for activating the vestibulo-muscular biostim array to compensate for a risk of a predicted fall.

Abstract: Described is system for tele-robotic operations over time-delayed communication links. Sensor data is acquired from at least one sensor for sensing surroundings of a robot having at least one robotic arm for manipulating an object. A three-dimensional model of the sensed surroundings is generated, and the sensor data is fit to the three-dimensional model. Using the three-dimensional model, a user demonstrates a movement path for the at least one robotic arm. A flow field representing the movement path is generated and combined with obstacle-repellent forces to provide force feedback to the user through a haptic device. The flow field comprises a set of parameters, and the set of parameters are transmitted to the robot to execute a movement of the at least one robotic arm for manipulating the object.

Abstract: A neural-Model Predictive Control (MPC) controller is described to control a dynamical system (i.e., “plant”). The MPC controller receives, in a base controller, a measurement of a current state of a plant and generates a control signal based on the measurement of the current state of the plant. A forward module receives the measurement of the current state of the plant and the control signal to generate a forward module prediction. A forward module corrector receives the measurement of the current state of the plant and the control signal from the base controller to generate an additive correction to the forward module prediction to generate a predicted plant state. Control sequences of length L of pairs of control signals and corresponding predicted plant states are generated until Ns control sequences have been generated. A next plant control signal is generated based on the Ns control sequences.

Abstract: A method for calibrating an articulable end effector of a robotic arm employing a digital camera includes commanding the end effector to achieve a plurality of poses. At each commanded end effector pose, an image of the end effector with the digital camera is captured and a scene point cloud including the end effector is generated based upon the captured image of the end effector. A synthetic point cloud including the end effector is generated based upon the commanded end effector pose, and a first position of the end effector is based upon the synthetic point cloud, and a second position of the end effector associated with the scene point cloud is determined. A position of the end effector is calibrated based upon the first position of the end effector and the second position of the end effector for the plurality of commanded end effector poses.

Abstract: Described is a robotic control device for manipulating a gripper-held tool. The device includes a robotic gripper having a plurality of tactile sensors. Each sensor generates tactile sensory data upon grasping an tool based on the interface between the tool and the corresponding tactile sensor. In operation, the device causes the gripper to grasp a tool and move the tool into contact with a surface. A control command is used to cause the gripper to perform a pseudo-random movement with the tool against the surface to generate tactile sensory data. A dimensionality reduction is performed on the tactile sensory data to generate a low-dimensional representation of the tactile sensory data, which is then associated with the control command to generate a sensory-motor mapping. A series of control commands can then be generated in a closed-loop based on the sensory-motor mapping to manipulate the tool against the surface.

Abstract: Described is a control system for stabilizing complex systems through self-adjustment. The complex system consists of agents or machines interacting with an environment and controlled by a controller. The control system includes a sensor configured to measure a state of the complex system and output the measured state of the complex system. A filter receives the measured state of the complex system, computes a variance in the measured state of the complex system over time, and outputs the computed variance. A regulator, which is connected with at least one controller, adjusts a control parameter in response to the computed variance received from the filter. The regulator is configured to regulate each controller's action on each agent or machine based on the control parameter in order to maintain stability of the complex system. In a desired aspect, the at least one control parameter comprises a set of additional input delays.

Abstract: A method of classification and segmentation of an image using modules on a computer system includes receiving a plurality of models having features suitable for classifying each pixel of the image into a respective one of a plurality of categories, using a classifier to provide a score for each pixel in the image for each category and using a segmenter to segment the image into image segments, wherein each image segment is a contiguous set of pixels having at least one common feature. For each image segment a set of average probabilities for each category is determined, and for each image segment, a most likely category to which the image segment belongs is determined by the maximum average probability resulting in a labeled segment image, which is used to identify any empty areas as incorrect holes. Then any empty areas that are identified as incorrect holes are filled.

Abstract: Described is a system for a system for autonomous robotic manipulation. The system is configured to receiving a selected task from a task file library. The task is associated with causing a robot end effector to perform an action with a particular item, such as picking up an item. The selected task is transformed into a state machine. Thereafter, the system executes the state machine and, in doing so, causes the robot end effector to perform the selected task.

Abstract: A robotic system includes an end-effector, an input device, and a controller. The input device is operable for collecting data defining a position and a velocity of a dynamic obstacle in an environment of the end-effector. The dynamic obstacle has an arbitrary shape. The controller is in communication with the end-effector and is programmed to execute a method and thereby receive a set of inputs via the input device, including the position and velocity of the dynamic obstacle. The controller computes a contour function defining the closest allowed distance and direction between the end-effector and the dynamic obstacle using the Gilbert-Johnson-Keerthi algorithm, and controls the end-effector via an output command to thereby avoid contact between the end-effector and the dynamic obstacle.

Abstract: A method for calibrating an articulable end effector of a robotic arm employing a digital camera includes commanding the end effector to achieve a plurality of poses. At each commanded end effector pose, an image of the end effector with the digital camera is captured and a scene point cloud including the end effector is generated based upon the captured image of the end effector. A synthetic point cloud including the end effector is generated based upon the commanded end effector pose, and a first position of the end effector is based upon the synthetic point cloud, and a second position of the end effector associated with the scene point cloud is determined. A position of the end effector is calibrated based upon the first position of the end effector and the second position of the end effector for the plurality of commanded end effector poses.

Abstract: A robotic system includes an end-effector and a control system. The control system includes a processor, a dynamical system module (DSM), and a velocity control module (VCM). Via execution of a method, the DSM processes inputs via a flow vector field and outputs a control velocity command. The inputs may include an actual position, desired goal position, and demonstrated reference path of the end-effector. The VCM receives an actual velocity of the end-effector and the control velocity command as inputs, and transmits a motor torque command to the end-effector as an output command. The control system employs a predetermined set of differential equations to generate a motion trajectory of the end-effector in real time that approximates the demonstrated reference path. The control system is also programmed to modify movement of the end-effector in real time via the VCM in response to perturbations of movement of the end-effector.

Abstract: Described is a system for controlling motion and constraint forces in a robotic system, the system infers constraints, at each time step, from a sensed state of robot/environment interactions. The inferred constraints are appended to known internal robot constraints to generate constrained dynamics. Properties associated with the constrained dynamics are determined provided to a controller. Inequality conditions associated with maintaining desired robot/environment interactions are also determined. A set of equality conditions based on the inequality conditions are then specified. The set of equality conditions are aggregated with any internal robot constraints to generate aggregated conditions that are provided to the controller. Joint torque commands are then generated for the robot based on the aggregated conditions and a specified task and null space motion command. Finally, the robot is actuated based on the joint torque commands.

Abstract: The present invention relates to a system for deep packet inspection and intrusion detection. The system uses a pattern matching module receiving as an input a data stream in a neural network. Neurons are activated such that when active, the neuron fires to all connecting output neurons to form a neuron spike, each neuron spike from the assigned neuron to a connecting output neuron having a delay. A delay is associated with each input character in the pattern, such that a position of each input character relative to an end of the pattern is stored in an alphabet-pattern-delay matrix (APDFM). An activation matrix (AM) is used to match each input character with a stored pattern to generate a similarity match and determine if the string of characters is the stored pattern.

Abstract: A robotic system includes an end-effector, an input device, and a controller. The input device is operable for collecting data defining a position and a velocity of a dynamic obstacle in an environment of the end-effector. The dynamic obstacle has an arbitrary shape. The controller is in communication with the end-effector and is programmed to execute a method and thereby receive a set of inputs via the input device, including the position and velocity of the dynamic obstacle. The controller computes a contour function defining the closest allowed distance and direction between the end-effector and the dynamic obstacle using the Gilbert-Johnson-Keerthi algorithm, and controls the end-effector via an output command to thereby avoid contact between the end-effector and the dynamic obstacle.

Abstract: Described is a system for searching a continuous data stream for exact matches with a priori stored data sequences. The system includes a neural network with an input and an output layer. The input layer has one neuron for each possible character or number in the data stream, and the output layer has one neuron for each stored pattern. Importantly, the delays of the connections from input to output layer are engineered to match the temporal occurrence of an input character within a stored sequence. Thus, if an input sequence has the proper time gaps between characters, matching a stored pattern, then the delayed neural signals result in a simultaneous activation at the receiving neuron, which indicates a detected pattern. For storing a pattern, only one connection for each pair of input character and output neuron has to be specified resulting in sparse coding and quick storage.

Abstract: Described is a system for localizing and controlling a tool tip with a robotic arm. The system receives three-dimensional (3D) visual information of a scene having a tool with a tool tip proximate a target. A depth disparity map is generated based on the 3D visual information. The depth disparity map is filtered by depth continuity to generate boundaries of the tool. The boundaries of the tool are initially loose to ensure that the tool is contained within the boundaries. Thereafter, using level set segmentation, the boundaries of the tool are shrunk and extracted. Ends of the tool tip are identified based on the extracted boundaries, which are then used to generate a pose of the tool tip.