Abstract: A system for detecting diseases among plants includes a mobile device such as a smartphone, table, or personal computer, and a mobile robot which sends photographs of diseased plants to the mobile device including a plant name, disease type and location of the plant. The mobile robot includes a power supply, housing unit, camera, navigation unit, GPS unit, wireless transmitter, and traction unit for mobility of the robot. A geared motor drives the traction unit, and is connected to a low-level controller for motion control. A high-level controller is connected to the camera, navigation unit, GPS unit, and programmed with a convolutional neural network (CNN) configured to process images taken by the camera and associate with each image a plant type, disease type, and location of where the image was taken. The CNN may be a you only look once (YOLO) model trained with images from a dataset.

Abstract: The robotic system for harvesting, and performing pre- and post-harvesting tasks of date palm trees includes three main parts: a truck platform, a scissor lift, and a multi-task system. The truck platform provides mobility for the whole system, while the scissor lift raises the multi-task system to the top of the palm tree. The multi-task system is fixed on top of the scissor lift and includes a large basin that catches fruits during harvesting. A circular track is fixed within the basin and includes a 2-degree-of-freedom (DOF) slider base mounted thereon. On top of the 2-DOF slider base is a robotic arm fitted with the proper tool (harvesting tool, maintenance tool or bagging system) for the task to be performed. The circular track has pivoting sections, and the basin has a flexible portion in the rear of the truck platform to allow the system to surround the palm tree.

Abstract: The tree harvesting tool is a tool adapted for attachment to a conventional lifter or lifting device for performing harvesting and pre-harvesting operations on a fruit tree. The tree harvesting tool includes a cylindrical shell, having an upper end and a lower end, at least one portion of the cylindrical shell defining a door. The cylindrical shell is adapted for encircling the trunk of the tree. A plurality of panels are pivotally secured to the lower end of the cylindrical shell to define an openable floor having a central opening for receiving the trunk of the tree. A circular track is mounted on the upper end of the cylindrical shell, such that a movable platform may be mounted thereon. A robotic arm is mounted on the movable platform for selectively operating and manipulating a tool for performing tree harvesting and pre-harvesting operations.

Abstract: The portable robot for two-way communication with the hearing-impaired provides for translation of audio into hand movements, hand shapes, hand orientations; the location of the hands around the body, the movements of the body and the head, facial expressions and lip movements to provide communication from a non-hearing-impaired individual(s) to hearing-impaired individual(s). The robot also translates hand movements, hand shapes, hand orientations; the location of the hands around the body, the movements of the body and the head, facial expressions and lip movements into audio to provide communication from hearing-impaired individual(s) to non-hearing-impaired individual(s). A communication system is provided to allow remote communication with non-hearing-impaired individual(s).

Abstract: A method and apparatus are described including determining if a received message is from a monitored peer, forwarding the received message to a coordinator responsive to said determination, determining if the received message is a request message, opening a port for the peer responsive to the second determination, forwarding the request message to a monitored peer via the opened port and forwarding a response message to the monitored peer responsive to the second determination.

Abstract: The present disclosure introduces a new technique for environmental recognition of audio input using feature selection. In one embodiment, audio data may be identified using feature selection. A plurality of audio descriptors may be ranked by calculating a Fisher's discriminant ratio for each audio descriptor. Next, a configurable number of highest ranking audio descriptors based on the Fisher's discriminant ratio of each audio descriptor are selected to obtain a selected feature set. The selected feature set is then applied to audio data. Other embodiments are also described.

Abstract: The present disclosure introduces a new technique for environmental recognition of audio input using feature selection. In one embodiment, audio data may be identified using feature selection. A plurality of audio descriptors may be ranked by calculating a Fisher's discriminant ratio for each audio descriptor. Next, a configurable number of highest ranking audio descriptors based on the Fisher's discriminant ratio of each audio descriptor are selected to obtain a selected feature set. The selected feature set is then applied to audio data. Other embodiments are also described.

Abstract: A method and apparatus are described including determining if a received message is from a monitored peer, forwarding the received message to a coordinator responsive to said determination, determining if the received message is a request message, opening a port for the peer responsive to the second determination, forwarding the request message to a monitored peer via the opened port and forwarding a response message to the monitored peer responsive to the second determination.