International Business Machines Corporation

Harry Roger Kolar

Eli M Dow

Michael L Passow

Matthew M Klawonn

Michael James Kelly

Michael R Kelly

Jessie Yu

Emanuele Ragnoli

Fearghal O'Donncha

Jeremiah Patrick Hayes

Frederick J Scheibl

Vasken Torossian

Noreen O'Brien

Quan G Cung

Kevin Eric Norsworthy

Julio Ortega

Ben Peter Yuhas

Embodiments of the invention are directed to computer-implemented methods, computer systems, and computer program products for remote monitoring of devices. The method includes causing a launch of an unmanned aerial vehicle (UAV). The UAV is instructed to fly a path that is optimized to read optical signals from a set of one or more devices. Data is received from the UAV, the data corresponding to the optical signals read from the set of one or more devices.

Remote sensor monitoring using LED transmission

An aspect includes aggregating a plurality of disparate datasets into a document store with semi-structured attributes, with each document in the document store specifying a geospatial location and a value of an environmental parameter at the geospatial location. The document store includes a plurality of documents specifying a plurality of different geo spatial locations and different environmental parameters, and the environmental parameters include a climatological parameter. Niche model layers are generated for the different environmental parameters at the geospatial locations. The niche models include a model layer for each of the different environmental parameters. A future geospatial location of a species is predicted based on environmental attributes of the species and contents of the niche model layers. The predicted future geospatial location of the species overlaid on a geographic map is output to a requestor.

Flexible framework for ecological niche modeling

An aspect includes aggregating a plurality of disparate datasets into a document store with semi-structured attributes that includes a plurality of documents specifying a plurality of different geo spatial locations and a plurality of different environmental parameters. Niche model layers are generated for the environmental parameters at the geospatial locations based on contents of the document store. The niche model layers include a model layer for each of the different environmental parameters. An additional niche model layer is created for a derived environmental parameter at the geospatial locations based at least in part on one of the previously generated niche model layers. A future geospatial location of a species is predicted based on environmental attributes of the species and contents of at least a subset of the niche model layers. The predicted future geospatial location of the species overlaid on a geographic map is output.

Inferring ecological niche model input layers and predicting a future geospatial location of a species

A system for mapping a thermocline in a body of fluid includes a thermocline detection and monitoring module, a persistent data storage module; and a plurality of distributed sensors, including intelligent sensors, connected with the at least one thermocline detection and monitoring module and the persistent data storage module by one or more control-level programming and communication methods. The thennocline detection and monitoring module can monitor the thermocline at sampling intervals to collect and fuse measurement data from the plurality of sensors to capture thermocline changes as events, correlate measurement data and events, store measurement data in the persistent data storage module along with previously acquired measurement data for comparison and tracking, characterize the thermocline as a function of spatial location, depth, and time, create and maintain reports that describe the thermocline characteristics, status, trends, and provide multimodal notifications of events to different users.

Intelligent/autonomous thermocline mapping and monitoring for marine and freshwater applications

An aspect includes aggregating a plurality of disparate datasets into a document store with semi-structured attributes, with each document in the document store specifying a geospatial location and a value of an environmental parameter at the geospatial location. The document store includes a plurality of documents specifying a plurality of different geo spatial locations and different environmental parameters, and the environmental parameters include a climatological parameter. Niche model layers are generated for the different environmental parameters at the geospatial locations. The niche models include a model layer for each of the different environmental parameters. A future geospatial location of a species is predicted based on environmental attributes of the species and contents of the niche model layers. The predicted future geospatial location of the species overlaid on a geographic map is output.

An aspect of the disclosure includes a method, a system and a computer program product. The method includes measuring a first data with a sensor, the sensor and operatively coupled computing resources having at least two power states, an operating power state and a low power state, the low power state consuming less energy than the operating power state. A first time period is determined for measuring a second data, the first time period being based at least in part on the first data. The sensor is operated, and operatively coupled computing resources at the low power state during the first time period, wherein no data measurements are acquired during the first time period. The sensor is changed to the operating power state at an expiration of the first time period.

Dynamic time sliced sensor sampling for reduced power consumption

A computer-implemented method includes calculating, by a processor of a computer system, a value for a turbulent flow parameter in a simulated model comprising an initial turbulent flow scheme for an advection-diffusion process; determining, by the processor of the computer system, that the value for the turbulent flow parameter has exceeded a predetermined threshold; conducting, by the processor of the computer system, a search to identify search results for other turbulent flow schemes which match the value of the turbulent flow parameter; determining, by the processor of the computer system, a quantified level of information confidence for the search results; annotating, by the processor of the computer system, the search results with the quantified level of information confidence; and selecting, by the processor of the computer system, a new turbulent flow scheme corresponding to the search result having the highest quantified level of information confidence.

Method and system for modelling turbulent flows in an advection-diffusion process

A computer-implemented method includes arranging two or more sensor nodes into a ring monitor. Each sensor node of the ring monitor is coupled to a sensor and cross-monitors at least one other sensor node in the ring monitor. A request is received to add a new sensor node. A spatial partitioning of a plane is generated, by a computer processor, representing locations of the new sensor node and the two or more sensor nodes in the ring monitor. The spatial partitioning includes a plurality of regions, each region corresponding to a sensor node. Two or more adjacent sensor nodes are selected from the two or more sensor nodes in the ring monitor, based at least in part on the spatial partitioning. The two or more adjacent sensor nodes are instructed to monitor the new sensor node, which is instructed to monitor each of the two or more adjacent sensor nodes.

Ring-based monitoring of sensor mesh networks

A solution for managing an asset, such as a movable asset, deployed to a fixed location, is provided. Location and/or power information is periodically acquired for the asset and assessed to determine whether the asset has moved from a desired physical area and/or the power is insufficient to continue with planned operations. In response to either condition, the operation of various devices on the asset can be adjusted to minimize an inferred risk caused by the asset moving outside of the physical area and/or maximize an inferred probability of recovering the asset.

Onboard management of movable asset for asset protection

Attributes of a data set to be employed in generating a predictive model are analyzed based on entropy, chi-square, or similar statistical measure. A target group of samples exhibiting one or more desired attributes is identified, then remaining attribute values for the target group are compared to corresponding attribute values for the whole sample population. A subset of all available attributes is then selected from those attributes which exhibit, when comparing attribute values of target group samples to attribute values for the whole sample population, the greatest relative difference or divergence. This subset is employed to generate the predictive model. Efficiency in generating the predictive model and the accuracy of the resulting predictive model is improved, since fewer attributes are employed and less computational resources are required.

Growing community of inventors

Scottsdale, AZ, United States of America

Harry Roger Kolar